A gas monitoring system for rock and soil tunneling operations

Abstract

This invention relates to the field of coal and gas outburst early warning and monitoring technology, and discloses a gas monitoring system for rock and soil tunneling operations, comprising: a rock mass space reconstruction module for initializing a three-dimensional voxel mesh model and constructing a transient cutting surface characterizing the physical unloading boundary based on real-time cutterhead pose data; a parameterized attribute evolution unit for determining initial damage based on the unloading gradient of the cutting surface and calculating anisotropic diffusion distribution along the normal vector to determine the permeability state; and a fracture network early warning module for identifying physical channels connecting the original stress equilibrium zone and the cutting surface through connectivity topology analysis of three-dimensional damage clusters and outputting early warning signals. This invention establishes a physical correlation between tunneling actions and gas migration, solves the problem of early warning failure caused by gas diffusion lag by characterizing the topological connectivity trend of rock mass structure failure, and enhances the reliability of disaster early warning under deep tunneling conditions.

Description

Technical Field

[0001] This invention relates to the field of coal and gas outburst early warning and monitoring technology, and more specifically, to a gas monitoring system for rock and soil tunneling operations. Background Technology

[0002] Currently, during the excavation of deep, high-stress strata, monitoring the risk of coal and gas outbursts is a crucial aspect of ensuring tunneling safety. Current technology typically involves deploying distributed gas sensors within the tunnel space to sample the gas concentration in the air and compare it with a preset threshold to output an alarm signal. This method, based on changes in environmental parameters, can provide basic monitoring under normal operating conditions. To enhance the monitoring capabilities, Chinese invention patent CN115506849B discloses a gas outburst monitoring and alarm device for underground tunneling faces. This device integrates displacement, wind speed and direction, and gas concentration sensors on a semi-circular platform and utilizes macroscopic phenomena such as coal wall bulging and local airflow fluctuations for early warning.

[0003] Coal and gas outbursts begin with the stress unloading of rock mass caused by mechanical cutting. This physical process is highly transient, and changes in gas concentration depend on the physical diffusion of gas in the rock mass and air. Its monitoring feedback lags behind the initial mechanical failure of the rock mass structure. Because the cutting entity of the tunneling machinery is continuously in a high-speed dynamic feeding state, the excavation boundary changes nonlinearly with the time series, resulting in a spatiotemporal mismatch between the scalar data collected by the sensors and the changes in the mechanical field inside the rock mass. To cope with the monitoring lag, the industry has tried to increase the density of sensor nodes or increase the sampling frequency. However, such linear optimization paths have failed to touch the geometric essence of the dynamic evolution of the excavation face. Existing systems usually set the excavation face as a static idealized boundary and lack the ability to calculate the Boolean interference of the mechanical entity on the volume stripping of the rock mass. This results in a geometric deviation between the mesh model in the topological space and the actual fracture fault, and it is impossible to accurately depict the connectivity trend of the fracture network in three-dimensional space.

[0004] Therefore, the technical problem to be solved by this invention is how to dynamically reconstruct the topological relationship between internal rock damage and flow field infiltration based on the real-time motion state of the tunneling machinery, and thereby identify the risk of geometric instability. Summary of the Invention

[0005] This invention provides a gas monitoring system for rock and soil tunneling operations, the system comprising: The rock mass space reconstruction module is used to initialize a three-dimensional voxel mesh model that represents the gas occurrence state of the rock mass in front of the tunneling machine, and to perform spatial interference calculation between the mechanical entity and the three-dimensional voxel mesh model based on the real-time pose data of the tunneling machine cutterhead, so as to peel off the voxel units of the corresponding volume from the three-dimensional voxel mesh model, thereby constructing a transient cutting surface that represents the physical unloading boundary. The parameterized attribute evolution unit is used to extract the three-dimensional normal vector of each boundary voxel element at the transient cutting surface, and determine the initial damage value of each boundary voxel element based on the instantaneous unloading gradient generated by the transient cutting surface. Using the direction of the three-dimensional normal vector as the transmission axis, anisotropic diffusion calculation is performed to determine the evolution distribution of local damage factor and permeability parameter within the three-dimensional voxel mesh model. The crack network early warning module is used to perform connectivity topology analysis on three-dimensional damage clusters defined by local damage factors and permeability parameters to identify physical channels connecting the original stress equilibrium zone and the transient cutting surface, and output gas outburst early warning signals based on the connectivity status of the physical channels.

[0006] Preferably, the parameterized attribute evolution unit constructs an unloading intensity distribution function based on the three-dimensional coordinate change rate of the transient cutting surface when performing anisotropic diffusion calculation; the parameterized attribute evolution unit is also used to obtain the displacement velocity of the transient cutting surface, and combined with the preset original rock mass ground pressure parameters, to calculate the stress release coefficient of each voxel unit in the three-dimensional voxel grid model in real time, so as to correct the calculation weight of the permeability parameter.

[0007] Preferably, the rock mass space reconstruction module uses a geometric Boolean subtraction algorithm to remove voxel elements that overlap with the spatial envelope of the tunnel boring machine cutterhead from the three-dimensional voxel mesh model when performing spatial interference calculations; the transient cutting surface is composed of a subset of newly generated boundary voxels after the execution of the geometric Boolean subtraction algorithm.

[0008] Preferably, the parameterized attribute evolution unit is also used to receive acoustic emission characteristic data at the transient cutting surface, and adjust the diffusion weight factor of the three-dimensional normal vector according to the energy centroid distribution with a frequency higher than 200Hz in the acoustic emission characteristic data, so that the diffusion pattern of the local damage factor inside the three-dimensional voxel mesh model is controlled by the dynamic occlusion boundary generated by mechanical cutting.

[0009] Preferably, the parameterized property evolution unit performs anisotropic diffusion calculation by constructing a permeability tensor matrix along the three-dimensional normal vector direction; the eigenvalues ​​of the permeability tensor matrix are positively correlated with the geometric normal of the transient cutting surface, which is used to eliminate the isotropic deviation generated by conventional spatial interpolation at the transient cutting surface.

[0010] Preferably, the fracture network early warning module defines each voxel in the three-dimensional voxel mesh model as a graph node when performing connectivity topology analysis, and establishes topological connection weights based on the permeability parameters between adjacent voxel units; the physical channel identification process includes: extracting the minimum weight path from the original stress equilibrium zone to the transient cutting surface.

[0011] Preferably, the original stress balance zone is defined by undisturbed rock strata in the area to be excavated ahead of the tunnel; the fracture network early warning module evaluates the evolution rate of the three-dimensional damage cluster by calculating the ratio of the geometric length of the physical channel to the topological hop number in the three-dimensional voxel mesh model, and uses this to define the danger level of the gas outburst early warning signal.

[0012] Preferably, the fracture network early warning module determines the trigger threshold of the gas outburst early warning signal by calculating the connectivity C. The formula for calculating the connectivity C is: , where C is the connectivity, and its value ranges from 0 to 1; The total volume of voxel units that meet the preset penetration threshold extracted from the physical channel by the fracture network early warning module; The total volume of voxels in the pre-set monitoring area in front of the tunnel boring machine cutterhead is set; when the connectivity C exceeds the pre-set instability critical threshold, a gas outburst warning signal is triggered.

[0013] Preferably, the system also includes a tunneling feedback control unit; the tunneling feedback control unit is used to receive gas outburst early warning signals and, based on the danger level in the gas outburst early warning signals, adjust the advance step distance and rotation speed of the tunneling machine cutterhead and the thrust pressure of the shield propulsion system in real time, so as to intervene in the release path of rock stress.

[0014] Preferably, the three-dimensional voxel mesh model is associated with a microsecond-level time stamp synchronization protocol; the rock mass space reconstruction module and the parameterized attribute evolution unit are both connected to the microsecond-level time stamp synchronization protocol to achieve time stamp alignment between the real-time pose data acquisition cycle and the three-dimensional voxel mesh model update cycle, so as to eliminate the physical perceived time delay between the mechanical unloading action and the gas response data stream.

[0015] The embodiments of the present invention have at least the following beneficial effects: 1. In rock and soil tunneling operations, the global clock synchronization node distributes nanosecond-level hardware time bases to the acoustic sensing unit and fluid sensing unit. Combined with the fluid transport delay coefficient dynamically calculated by the adaptive phase compensation unit, a strong causal mapping between the acoustic emission pulse sequence of rock mass fracture and the in-situ gas transient concentration sequence is established in the time domain. This dynamic phase alignment mechanism based on the real-time motion parameters of the cutting module eliminates the asynchronous signal error caused by the high-speed displacement of the tunneling head and the disturbance of the complex two-phase flow field. This enables the system to directly capture the instantaneous gas desorption response triggered by mechanical unloading, thereby accurately extracting the prominent incubation critical features with pre-existing attributes from the intense background noise generated by the tunneling machine cutting.

[0016] 2. The computer-aided 3D voxel reconstruction engine has a pre-built truncation module that represents the 3D boundary entity model. By performing temporal trajectory sweeping and continuous 3D Boolean difference set calculations in memory space, it dynamically peels off voxel units that spatially interfere with the mechanical entity. This allows for the real-time construction of a transient cutting surface representing the physical peeling boundary at the bottom layer of the computing architecture. This mechanism translates the underground rock excavation process into a geometric evolution of entity interference determined in memory, effectively avoiding the boundary condition modeling distortion caused by treating the excavation surface as a static envelope in the time domain in traditional schemes. It ensures the absolute geometric rigidity of the image position of all perceived data in the 3D coordinate system.

[0017] 3. The parametric topology algorithm extracts the three-dimensional normal gradient vector of each boundary voxel element on the transient cutting surface and uses it as the starting point for scalar source propagation. This drives the local damage geometric parameters and fluid permeability properties to diffuse anisotropically along the vector direction. This parameter propagation mode based on the dynamic cutting surface geometry orientation breaks the isotropic deviation of conventional spatial interpolation algorithms under complex mechanical obstruction and flow field distortion. It realizes the high-fidelity directional penetration of acoustic and gas coupling parameters into the micro-fracture network inside the rock mass, so that the reconstructed three-dimensional voxel model can accurately depict the real extension and connection morphology of the fracture network in three-dimensional space. Attached Figure Description

[0018] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings, in which several embodiments of the invention are illustrated by way of example and not limitation, wherein: Figure 1 This is the main flowchart of the gas monitoring and outburst early warning system for rock and soil tunneling of this invention; Figure 2 This is a schematic diagram illustrating the principle of gas monitoring multi-parameter evolution and fracture network topology determination in this invention. Detailed Implementation

[0019] The principles and spirit of the present invention will now be described with reference to several exemplary embodiments in conjunction with the accompanying drawings. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0020] A gas monitoring system for rock and soil tunneling operations, the system comprising: The rock mass space reconstruction module is used to initialize a three-dimensional voxel mesh model that represents the gas occurrence state of the rock mass in front of the tunneling machine, and to perform spatial interference calculation between the mechanical entity and the three-dimensional voxel mesh model based on the real-time pose data of the tunneling machine cutterhead, so as to peel off the voxel units of the corresponding volume from the three-dimensional voxel mesh model, thereby constructing a transient cutting surface that represents the physical unloading boundary. The parameterized attribute evolution unit is used to extract the three-dimensional normal vector of each boundary voxel element at the transient cutting surface, and determine the initial damage value of each boundary voxel element based on the instantaneous unloading gradient generated by the transient cutting surface. Using the direction of the three-dimensional normal vector as the transmission axis, anisotropic diffusion calculation is performed to determine the evolution distribution of local damage factor and permeability parameter within the three-dimensional voxel mesh model. The crack network early warning module is used to perform connectivity topology analysis on three-dimensional damage clusters defined by local damage factors and permeability parameters to identify physical channels connecting the original stress equilibrium zone and the transient cutting surface, and output gas outburst early warning signals based on the connectivity status of the physical channels.

[0021] Preferably, the parameterized attribute evolution unit constructs an unloading intensity distribution function based on the three-dimensional coordinate change rate of the transient cutting surface when performing anisotropic diffusion calculation; the parameterized attribute evolution unit is also used to obtain the displacement velocity of the transient cutting surface, and combined with the preset original rock mass ground pressure parameters, to calculate the stress release coefficient of each voxel unit in the three-dimensional voxel grid model in real time, so as to correct the calculation weight of the permeability parameter.

[0022] Preferably, the rock mass space reconstruction module uses a geometric Boolean subtraction algorithm to remove voxel elements that overlap with the spatial envelope of the tunnel boring machine cutterhead from the three-dimensional voxel mesh model when performing spatial interference calculations; the transient cutting surface is composed of a subset of newly generated boundary voxels after the execution of the geometric Boolean subtraction algorithm.

[0023] Preferably, the parameterized attribute evolution unit is also used to receive acoustic emission characteristic data at the transient cutting surface, and adjust the diffusion weight factor of the three-dimensional normal vector according to the energy centroid distribution with a frequency higher than 200Hz in the acoustic emission characteristic data, so that the diffusion pattern of the local damage factor inside the three-dimensional voxel mesh model is controlled by the dynamic occlusion boundary generated by mechanical cutting.

[0024] Preferably, the parameterized property evolution unit performs anisotropic diffusion calculation by constructing a permeability tensor matrix along the three-dimensional normal vector direction; the eigenvalues ​​of the permeability tensor matrix are positively correlated with the geometric normal of the transient cutting surface, which is used to eliminate the isotropic deviation generated by conventional spatial interpolation at the transient cutting surface.

[0025] Preferably, the fracture network early warning module defines each voxel in the three-dimensional voxel mesh model as a graph node when performing connectivity topology analysis, and establishes topological connection weights based on the permeability parameters between adjacent voxel units; the physical channel identification process includes: extracting the minimum weight path from the original stress equilibrium zone to the transient cutting surface.

[0026] Preferably, the original stress balance zone is defined by undisturbed rock strata in the area to be excavated ahead of the tunnel; the fracture network early warning module evaluates the evolution rate of the three-dimensional damage cluster by calculating the ratio of the geometric length of the physical channel to the topological hop number in the three-dimensional voxel mesh model, and uses this to define the danger level of the gas outburst early warning signal.

[0027] Preferably, the fracture network early warning module determines the trigger threshold of the gas outburst early warning signal by calculating the connectivity C. The formula for calculating the connectivity C is: , where C is the connectivity, and its value ranges from 0 to 1; The total volume of voxel units that meet the preset penetration threshold extracted from the physical channel by the fracture network early warning module; The total volume of voxels in the pre-set monitoring area in front of the tunnel boring machine cutterhead is set; when the connectivity C exceeds the pre-set instability critical threshold, a gas outburst warning signal is triggered.

[0028] Preferably, the system also includes a tunneling feedback control unit; the tunneling feedback control unit is used to receive gas outburst early warning signals and, based on the danger level in the gas outburst early warning signals, adjust the advance step distance and rotation speed of the tunneling machine cutterhead and the thrust pressure of the shield propulsion system in real time, so as to intervene in the release path of rock stress.

[0029] Preferably, the three-dimensional voxel mesh model is associated with a microsecond-level time stamp synchronization protocol; the rock mass space reconstruction module and the parameterized attribute evolution unit are both connected to the microsecond-level time stamp synchronization protocol to achieve time stamp alignment between the real-time pose data acquisition cycle and the three-dimensional voxel mesh model update cycle, so as to eliminate the physical perceived time delay between the mechanical unloading action and the gas response data stream.

[0030] Example 1: When the system faces the mechanized tunneling of deep high-pressure rock strata, the undisturbed rock strata in the area to be excavated ahead contain high-stress gas fluid. The high-speed rotation of the tunneling machine cutterhead and the thrust feed continuously apply mechanical cutting to the rock strata ahead. This transient unloading action disrupts the original stress balance and induces initial damage to the internal structure of the rock mass and gas desorption. The desorbed gas needs to be transported over a long distance through the pore network of the newly formed rock strata to reach the tunnel space. The traditional environmental concentration exceeding limit comparison mechanism relies on free gas molecules in the tunnel space, which has an inherent physical lag. This leads to a spatiotemporal mismatch between the monitoring data and the mechanical fracture faults of the underlying rock mass, making it impossible to provide a definite basis for intervention in the early stage of the outburst disaster. The rock mass spatial reconstruction module initializes a three-dimensional voxel mesh model representing the gas occurrence state of the rock mass ahead of the tunneling. It then uses a microsecond-level time stamp synchronization protocol to acquire real-time pose data of the tunneling machine cutterhead. A geometric Boolean subtraction algorithm is used to remove voxel elements from the three-dimensional voxel mesh model that overlap with the spatial envelope of the tunneling machine cutterhead. This entity interference calculation process outputs a transient cutting surface composed of a subset of newly formed boundary voxels. The parameterized attribute evolution unit extracts the three-dimensional normal vector of each boundary voxel element at the transient cutting surface, transforming the entity evolution boundary of the excavation face into a geometric gradient frame with a single direction, thus eliminating the perceived time delay between the mechanical unloading action and the gas response data stream.

[0031] The parameterized attribute evolution unit constructs an unloading intensity distribution function based on the three-dimensional coordinate change rate of the transient cutting surface, determines the initial damage value of each boundary voxel unit, synchronously receives acoustic emission characteristic data at the transient cutting surface, extracts the energy centroid distribution with frequencies higher than 200Hz, adjusts the diffusion weight factor of the three-dimensional normal vector based on the acoustic characteristics of this frequency band, constructs a permeability tensor matrix with the direction of the three-dimensional normal vector as the transmission axis, and performs anisotropic diffusion calculation. This directional transmission mechanism eliminates the isotropic bias of conventional interpolation algorithms, drives the evolution of local damage factors and permeability parameters within the three-dimensional voxel mesh model, and reconstructs the mechanical rock-breaking unloading event as the dynamic evolution characteristics of the physical permeability field topology in three-dimensional space. The fracture network early warning module defines each voxel unit in the three-dimensional damage cluster defined by the local damage factors and permeability parameters as a graph node, establishes topological connection weights based on the permeability parameters between adjacent voxel units, performs connectivity topology analysis, extracts the minimum weight path from the original stress equilibrium zone to the transient cutting surface, identifies the connected physical channels, and calculates the connectivity C of the physical channels according to the formula. The triggering logic for the gas outburst early warning signal is determined; where C is a dimensionless scalar parameter characterizing the topological connectivity state of the three-dimensional damage cluster. This refers to the total volume of voxel units that meet the preset penetration threshold extracted from the physical channel by the fracture network early warning module. For the total volume of voxels in the pre-set monitoring area in front of the tunnel boring machine cutterhead, when the connectivity C is greater than the preset instability critical threshold, the fracture network early warning module calculates the ratio of the geometric length of the physical channel to the number of topological hops, defines the danger level, and outputs a gas outburst early warning signal. The tunneling feedback control unit receives the gas outburst early warning signal and adjusts the advance step distance, rotation speed, and thrust pressure of the tunnel boring machine cutterhead in real time according to the danger level, intervenes in the release path of rock mass stress, and blocks the evolution of gas disaster.

[0032] Example 2: When the system faces deep, high-stress rock and soil tunneling conditions, the experiment relies on a true three-dimensional gas-bearing coal-bearing rock dynamic disaster simulation platform to obtain physical verification data. This simulation platform's servo loading unit has a three-dimensional principal stress range of 100 MPa and a gas injection pressure limit of 5 MPa. It is internally equipped with a scaled-down tunneling machine cutterhead and an acoustic emission sensor array with a sampling rate of 1 MHz. The experiment continuously mixes 85 dB of mechanical vibration Gaussian white noise and 50 Hz power frequency interference harmonics into the raw acoustic signal of the acoustic emission sensor array, generating an initial test environment containing background disturbances. The rock mass space reconstruction module initializes the three-dimensional rock mass space in the initialization phase. When creating a voxel mesh model, the voxel side length parameter is selected. This parameter is controlled by the dominant frequency wavelength limit to maintain a balance between spatial reconstruction resolution and the computational load of the three-dimensional interferometry. When the dominant frequency wavelength index above 200Hz captured by the acoustic emission sensor array reaches a predetermined value, the upper limit of the voxel side length parameter is set to one-quarter of the dominant frequency wavelength. Based on the measured longitudinal wave velocity of 2500m / s in the rock mass and the dominant frequency of 500Hz, a dominant frequency wavelength of 5m is calculated. Therefore, the upper limit of the voxel side length parameter is defined as 1.25m. This experimental group selected 0.5m as the working reference parameter for the three-dimensional voxel mesh model. satisfy ,in Let be the side length of a voxel in a three-dimensional voxel mesh model, and λ be the dominant frequency wavelength of the acoustic emission signal generated by the tunnel boring machine cutterhead cutting the rock. Furthermore, λ is based on... Confirmed, among which In order to measure the longitudinal wave velocity of the rock mass in advance, To enable the acoustic emission sensor array to capture the centroid frequency of the acoustic emission signal energy in real time, during spatial interferometry calculations, the rock mass space reconstruction module obtains the three-dimensional spatial coordinates (x, y, z) of the tunnel boring machine cutterhead and its axial propulsion vector in real time, based on feedback from the control system. An envelope model with geometric dimensions equivalent to the tunnel boring machine cutterhead is generated in memory. A geometric Boolean subtraction operator is called to traverse and compare the index addresses of each voxel element in the envelope model's coordinate space with those in the three-dimensional voxel mesh model. Voxel elements with their center coordinates located inside the envelope model are marked as excavated and removed from the current calculation array. A transient cutting surface is constructed from a subset of newly formed boundary voxels.

[0033] The scaled-down tunneling machine cutterhead feeds and cuts gas-bearing coal-rock mass with a thrust of 500kN. The parametric attribute evolution unit extracts the original acoustic emission signal containing the aforementioned mechanical vibration noise and power frequency interference harmonics. The initial signal-to-noise ratio of this original acoustic emission signal is 3.2dB. The parametric attribute evolution unit extracts acoustic features with frequencies higher than 200Hz and separates the energy centroid distribution, filtering out low-frequency mechanical noise components. The output energy centroid distribution data has a signal-to-noise ratio of 18.5dB. The parametric attribute evolution unit performs a fast Fourier transform on the original acoustic emission signal acquired at the transient cutting surface, extracts the spectral features in the frequency band between 200Hz and 500Hz, and calculates the energy centroid in this frequency band. 3D normal vector diffusion weighting factor satisfy ,in For diffusion weighting factors, To extract the energy center of gravity in real time, To obtain the baseline energy centroid constant under a pre-set gas-bearing coal and rock calibration environment, the parameterized attribute evolution unit will calculate the diffusion weighting factor. The permeability tensor matrix is ​​stored in the diagonal components, so that the diffusion rate of the local damage factor along the three-dimensional normal vector direction during the evolution process is controlled by the acoustic characteristics excited by mechanical cutting. A non-uniform damage evolution field positively correlated with the mechanical unloading intensity is formed inside the three-dimensional voxel mesh model. The parameterized attribute evolution unit adjusts the diffusion weight factor of the three-dimensional normal vector according to the energy centroid distribution, constructs the permeability tensor matrix along the three-dimensional normal vector direction and performs anisotropic diffusion calculation, and outputs dynamically updated permeability parameters.

[0034] The experiment established a control group and compared it with the sample group of the present invention. The control group used an isotropic spatial distance interpolation algorithm to calculate the permeability parameter, while the sample group of the present invention used an anisotropic diffusion calculation that included the weighting adjustment of the three-dimensional normal vector and the centroid of acoustic emission energy. When the data acquisition period reached 12.5s, the three-dimensional damage cluster in the control group showed spherical uniform expansion with a connectivity C of 0.15. At this time, the peak gas emission measured by the simulation platform had occurred, and the data of the control group showed a spatiotemporal deviation from the physical rupture fault. The permeability parameter in the sample group of the present invention was controlled by the dynamic boundary normal of mechanical cutting, with a connectivity C of 0.42. The evolution trajectory of this connected topology matched the physical evolution node of gas desorption. The data showed that the geometric spatial normal and the acoustic energy centroid had a synergistic output result that offset the interpolation bias error. The fracture network early warning module was established using the formula... The triggering conditions for gas outburst warning signals were determined. The experiment established a gradient array for the instability critical threshold, including a lower limit test point of 0.20, a median test point of 0.35, an upper limit test point of 0.50, and an out-of-range test point of 0.60. When the instability critical threshold was set to 0.20, the system output a warning signal 5.2 seconds after cutting started, with a false alarm rate of 41.5% due to surface spalling microcracks. When set to the median value of 0.35, the system captured the topological continuity characteristics of the main control fracture network at 6.1 seconds. When set to the upper limit of 0.50, the warning output time was 7.8 seconds. When set to the out-of-range value of 0.60, the warning trigger time was delayed to 8.7 seconds, lagging behind the gas pressure abrupt change inflection point recorded by the simulation platform at 6.3 seconds. Data comparison defined 0.35 to 0.50 as the working range for the instability critical threshold, and the gas outburst warning signal instability critical threshold... The following calibration was used to determine the connectivity C value at the moment of physical fracturing abruptly occurring in the rock mass, using a true 3D gas-bearing coal-bearing rock dynamic disaster simulation platform during the simulated tunneling machine cutterhead propulsion experiment; the experiment was repeated at least 5 times to obtain a set of connectivity sample values ​​corresponding to the physical fracturing nodes, and the arithmetic mean of the set of sample values ​​was calculated. ; the arithmetic mean Set as the instability critical threshold The value is pre-stored in the logic identification circuit of the fracture network early warning module; during actual tunneling operations, the fracture network early warning module will calculate the connectivity C and the instability critical threshold in real time. A magnitude comparison is performed; when the connectivity C is greater than or equal to the instability critical threshold... At that time, the fracture network early warning module blocks the normal monitoring state and triggers the output of the gas outburst early warning signal. The fracture network early warning module quantifies the connection status of the physical channel from the original stress balance zone to the transient cutting surface based on the connectivity C parameter defined in the working range. This quantification feature is independent of the environmental gas free molecule concentration parameter, matches the fracture progress of the underlying rock mass, and outputs the execution command for disaster prevention and control of mechanized tunneling in soil and rock.

[0035] Example 3: When the system faces continuous excavation of deep, high-stress rock mass, after the geometric Boolean subtraction algorithm is used to eliminate voxel arrays, the boundary of the remaining voxel array in the three-dimensional voxel mesh model presents a discrete step-like topological shape. Extracting this discrete boundary causes the normal vector calculation to diverge. Relying solely on coordinate parameters cannot quantify the true amplitude of the rock mass mechanical response, resulting in the calculation weight of the permeability parameter lacking the underlying physical boundary basis, causing the anisotropic diffusion distribution model to become disconnected from the actual fracture fault. The parameterized attribute evolution unit locks the transient cutting surface composed of a subset of newly formed boundary voxels. For each boundary voxel unit on this surface, the spatial neighborhood matrix containing 26 adjacent voxels is retrieved. Based on the distribution of the solid rock voxels and the excavated empty voxels in the neighborhood matrix, the three-dimensional occupancy gradient is calculated. The three-dimensional occupancy gradient is orthogonally decomposed along the coordinate axis and normalized to output a smooth three-dimensional normal vector.

[0036] The parameterized attribute evolution unit extracts the transient cutting surface displacement velocity v fed back by the tunnel boring machine cutterhead propulsion system and reads the original rock mass ground pressure parameters pre-recorded in the exploration. And the rock mass acoustic impedance Z, according to the formula The stress release coefficient K of each boundary voxel element within the three-dimensional voxel mesh model is calculated in real time; where K is a dimensionless scalar coefficient, Z is the product of rock mass density and P-wave velocity, and v is the displacement velocity. The original ground pressure parameters of the rock mass are given. The dynamic unloading stress on the numerator of the right side of the equation has dimensions of Pa, which is equivalent to the dimension of the denominator. This establishes the dimensional consistency of the coefficient calculation. The parameterized attribute evolution unit corrects the calculation weight of the permeability parameter based on the stress release coefficient K. When the displacement velocity v increases sharply, causing the stress release coefficient K to increase, the calculation weight simultaneously amplifies the anisotropic principal axis component of the normal permeability tensor, driving the diffusion trajectory of the local damage factor to obey the real unloading law of rock dynamics. This transforms the spatial interference at the geometric level into a basis for reconstructing the gas permeability field that matches the laws of physical mechanics.

[0037] Example 4: When the system faces the deployment of a new working face in a heterogeneous geological structure zone, the initial parameter deviation caused by local structural stress anomalies when the wide-area geological exploration parameters are substituted; before the tunnel boring machine cutterhead cuts, the rock mass space reconstruction module starts the on-site pre-baseline calibration program, drives the acoustic emission sensor array to emit broadband calibration sound waves with a frequency range of 200Hz to 500Hz into the undisturbed rock mass in front of the tunnel boring machine cutterhead, and collects the corresponding echo signal for a duration of up to 10s; the parameterized attribute evolution unit extracts the energy attenuation envelope of the echo signal to calculate the local rock mass acoustic impedance of each coordinate node in the three-dimensional voxel mesh model. Simultaneously calculate the expected value and standard deviation of the natural background permeability parameter of the rock mass within a 10-second sampling window. Define the preset permeability threshold used to identify physical channels based on the sum of the expected value and three times the standard deviation. Static background bias data of geological heterogeneous injection were extracted.

[0038] After the on-site baseline calibration procedure is completed, the fracture network early warning module will preset the permeability threshold. Loaded into the connectivity topology analysis logic as a rigid filter boundary for node connectivity; during the unloading of rock mass induced by mechanical cutting, the parameterized attribute evolution unit forces the dynamically calculated permeability parameters of each boundary voxel unit to be compared with the preset permeability threshold. One by one, the permeability parameter is compared, and when it is greater than the preset permeability threshold... At that time, the corresponding voxel unit is marked as a positive mutation node and included in the formula characterizing the three-dimensional damage cluster. Intermolecular term The volume accumulation pool; the differential comparison process removes the illusion of permeability fluctuations induced by simple elastic compression around the transient cutting surface, and converges the identified physical channels to the gas desorption and transport network dominated by irreversible plastic damage. The quantitative characteristics stably reflect the true evolution state of mechanical fracture of the underlying rock mass, and output a disaster prevention and control parameter sequence for mechanized tunneling in soil and rock that is adapted to complex geological disturbances.

[0039] Example 5: When the system faces long-distance continuous rock and soil tunneling, the initial stress field boundary of the rock mass ahead dynamically migrates forward as the tunnel boring machine cutterhead continues to cut. In the three-dimensional voxel mesh model, the original stress equilibrium zone is defined using absolute static coordinates, causing the search space of the topology pathfinding algorithm to continuously increase with the excavation distance and the calculation of connectivity data to diverge. The rock mass space reconstruction module executes a spatial sliding window tracking process during the initialization phase, establishing a follow-up local reference system with the center of the tunnel boring machine cutterhead as the reference coordinate origin, and setting the detection depth parameter along the cutting direction. The geometry of the original stress equilibrium zone is dynamically defined; among them, To characterize the spatial axial distance parameter of the effective monitoring domain depth, the test control unit records the background acoustic signal reflected by the rock mass ahead under constant thrust. The spatial first derivative of the acoustic signal envelope amplitude is extracted, and the spatial axial distance at which it decays to near zero is determined as the detection depth parameter. The working values ​​are used to define the effective monitoring domain that accompanies the synchronous advancement of mechanical cutting in a three-dimensional voxel mesh model.

[0040] After determining the effective monitoring domain, the fracture network early warning module extracts the permeability parameters of adjacent voxel units within the effective monitoring domain. It then generates the topological connection weights of graph nodes by calculating the reciprocal of the arithmetic mean of the permeability parameters of two adjacent voxel units. This reciprocal calculation step transforms high-permeability physical fracture areas into low-weight network paths. The fracture network early warning module then calls the shortest path optimization algorithm to find paths located at the depth of the detection field. The set of voxel units at a given location serves as the starting search surface, and the subset of voxel units contained in the transient cutting surface serves as the target termination surface. The topological connection weights on the physical evolution path are iteratively accumulated to select the connected sequence with the smallest cumulative value. This node search step removes data nodes in irrelevant static regions of the model, restricts the identification boundary of the physical channel to the active space affected by mechanical unloading disturbance, reduces the computational load of graph node iteration, and provides a basis for calculating connectivity parameters under long-distance dynamic excavation conditions.

[0041] The above description is only a few preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, technical solutions formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention.

Claims

1. A gas monitoring system for rock and soil tunneling operations, characterized in that the system... include: The rock mass space reconstruction module is used to initialize a three-dimensional voxel mesh model that represents the gas occurrence state of the rock mass in front of the tunneling machine, and to perform spatial interference calculation between the mechanical entity and the three-dimensional voxel mesh model based on the real-time pose data of the tunneling machine cutterhead, so as to peel off the voxel units of the corresponding volume from the three-dimensional voxel mesh model, thereby constructing a transient cutting surface that represents the physical unloading boundary. The parameterized attribute evolution unit is used to extract the three-dimensional normal vector of each boundary voxel element at the transient cutting surface, and determine the initial damage value of each boundary voxel element based on the instantaneous unloading gradient generated by the transient cutting surface. Using the direction of the three-dimensional normal vector as the transmission axis, anisotropic diffusion calculation is performed to determine the evolution distribution of local damage factor and permeability parameter within the three-dimensional voxel mesh model. The crack network early warning module is used to perform connectivity topology analysis on three-dimensional damage clusters defined by local damage factors and permeability parameters to identify physical channels connecting the original stress equilibrium zone and the transient cutting surface, and output gas outburst early warning signals based on the connectivity status of the physical channels.

2. The gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The parameterized attribute evolution unit constructs the unloading intensity distribution function based on the three-dimensional coordinate change rate of the transient cutting surface when performing anisotropic diffusion calculations. The parameterized attribute evolution unit is also used to obtain the displacement velocity of the transient cutting surface and, in combination with the preset original rock mass ground pressure parameters, to calculate the stress release coefficient of each voxel element in the three-dimensional voxel mesh model in real time, so as to correct the calculation weight of the permeability parameter.

3. The gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The rock mass spatial reconstruction module uses a geometric Boolean subtraction algorithm to remove voxel elements that overlap with the spatial envelope of the tunnel boring machine cutterhead from the three-dimensional voxel mesh model when performing spatial interference calculations; the transient cutting surface is composed of a subset of newly generated boundary voxels after the execution of the geometric Boolean subtraction algorithm.

4. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The parameterized attribute evolution unit is also used to receive acoustic emission characteristic data at the transient cutting surface, and adjust the diffusion weight factor of the three-dimensional normal vector according to the energy centroid distribution with a frequency higher than 200Hz in the acoustic emission characteristic data, so that the diffusion pattern of the local damage factor inside the three-dimensional voxel mesh model is controlled by the dynamic occlusion boundary generated by mechanical cutting.

5. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The parameterized property evolution unit performs anisotropic diffusion calculations by constructing a permeability tensor matrix along the three-dimensional normal vector direction. The eigenvalues ​​of the permeability tensor matrix are positively correlated with the geometric normal of the transient cutting surface, which is used to eliminate the isotropic deviation caused by conventional spatial interpolation at the transient cutting surface.

6. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The crack network early warning module defines each voxel in the three-dimensional voxel mesh model as a graph node when performing connectivity topology analysis, and establishes topological connection weights based on the permeability parameters between adjacent voxel units; the physical channel identification process includes: extracting the minimum weight path from the original stress equilibrium zone to the transient cutting surface.

7. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The original stress equilibrium zone is defined by the undisturbed rock strata pre-set in the area to be excavated ahead of the tunnel; The crack network early warning module assesses the evolution rate of three-dimensional damage clusters by calculating the ratio of the geometric length of the physical channel to the number of topological hops in the three-dimensional voxel mesh model, and uses this to define the danger level of the gas outburst early warning signal.

8. A gas monitoring system for rock and soil tunneling operations according to claim 7, characterized in that, The fracture network early warning module determines the trigger threshold for a gas outburst early warning signal by calculating the connectivity C. The formula for calculating the connectivity C is: , where C is the connectivity, and its value ranges from 0 to 1; The total volume of voxel units that meet the preset penetration threshold extracted from the physical channel by the fracture network early warning module; The total volume of voxels in the pre-set monitoring area in front of the tunnel boring machine cutterhead is set; when the connectivity C exceeds the pre-set instability critical threshold, a gas outburst warning signal is triggered.

9. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The system also includes a tunneling feedback control unit; the tunneling feedback control unit is used to receive gas outburst early warning signals and, based on the danger level in the gas outburst early warning signals, adjust the advance step distance and rotation speed of the tunneling machine cutterhead and the thrust pressure of the shield propulsion system in real time to intervene in the release path of rock stress.

10. A gas monitoring system for rock and soil tunneling operations according to claim 1, characterized in that, The three-dimensional voxel mesh model is associated with a microsecond-level time stamp synchronization protocol; the rock mass spatial reconstruction module and the parameterized attribute evolution unit are both connected to the microsecond-level time stamp synchronization protocol to achieve time stamp alignment between the real-time pose data acquisition cycle and the three-dimensional voxel mesh model update cycle, so as to eliminate the physical perceived time delay between mechanical unloading action and gas response data stream.

Images