Data monitoring and analysis method and system for shield tunneling in submarine karst strata

Abstract

The present invention discloses a method and system for monitoring and analyzing shield tunneling data in submarine karst formations. The method comprises: synchronously collecting shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data; inverting the shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data based on seepage-stress coupling constraints, generating a hydraulic connectivity strength index through fracture network permeability tensor inversion; determining the risk level of filling migration based on the dynamic mutation characteristics of pore water pressure; and generating a shield disturbance entropy value by integrating the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics. The present invention generates a hydraulic connectivity index through seepage-stress coupling inversion and dynamically optimizes the grouting path and rheological parameters based on the fracture probability field, achieving coordinated shield tunneling, seepage mutation, and grouting control, fundamentally resolving the problem of disconnected control.

Description

Technical Field

[0001] The present application relates to the technical field of tunnel construction monitoring, and in particular to a method and system for monitoring and analyzing shield tunneling data in submarine karst strata. Background Art

[0002] Cross-sea tunnel projects are increasingly extending into high-water-pressure, highly karst formations, and shield tunneling is facing unprecedented risks of sudden surge disasters. Shield tunneling in submarine karst formations faces two core challenges:

[0003] First, the collaborative prediction of multi-source heterogeneous data fails. Traditional methods are unable to integrate multi-dimensional heterogeneous data such as mechanical construction parameters (cutter head torque, propulsion force), formation response (acoustic emission signal) and hydrological dynamics (seawater osmotic pressure), resulting in insufficient accuracy in karst outburst risk prediction. Existing technologies only rely on single pore water pressure monitoring, ignoring the dynamic coupling mechanism between shield mechanical disturbance and fracture network seepage, resulting in a high false alarm rate.

[0004] Second, risk prevention and control are disconnected from construction control. Conventional grouting decisions lag behind the formation mutation process. Grouting parameters (pressure, rheology) and propulsion parameters (thrust, speed) cannot respond to the dynamic evolution of hydraulic connectivity strength and filling migration risk in real time. Especially when there is a gradient mutation in the karst pipeline rupture probability field, there is a lack of a spatial position-driven graded early warning mechanism. Summary of the Invention

[0005] In order to solve the above problems, an embodiment of the present invention provides a method for monitoring and analyzing shield tunneling data in submarine karst strata, the method comprising:

[0006] Simultaneously collect shield construction machinery parameters, karst formation acoustic emission signals and seawater osmotic pressure data;

[0007] Based on the seepage-stress coupling constraint, the parameters of shield construction machinery, karst formation acoustic emission signals, and seawater osmotic pressure data were used to generate a hydraulic connectivity strength index through fracture network permeability tensor inversion. The risk level of filling migration was determined based on the dynamic mutation characteristics of pore water pressure, and the shield disturbance entropy value was generated by integrating the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics.

[0008] The hydraulic connectivity strength index, filling material migration risk level, and shield disturbance entropy value are input into the spatiotemporal graph convolutional network. The shield thrust-karst displacement mapping is constructed through the mechanical response subnet. The karst fluid velocity is predicted through the seepage correlation subnet. Finally, the karst pipeline rupture probability field and shield control compensation are output.

[0009] According to the distribution of gradient mutation zones of the rupture probability field and their spatial relationship with the shield cutterhead, a graded warning signal is generated.

[0010] Based on the shield control compensation amount and the risk level of filling migration, the grouting pressure distribution and advancement parameters are dynamically optimized.

[0011] Furthermore, the reconstruction method of the hydraulic connectivity strength index includes:

[0012] The three-dimensional resistivity distribution of the karst formation is coupled with the pore water pressure gradient vector field, and the hydraulic connectivity strength index is generated using the permeability tensor inversion algorithm. The permeability tensor inversion algorithm is constructed based on Biot's seepage-stress theory, and its output value is used to characterize the fluid conductivity efficiency of the karst fracture network.

[0013] Furthermore, the method for determining the risk level of filling material migration includes:

[0014] The nonlinear dynamic phase space of the pore water pressure time series is reconstructed, and the sudden change in the dynamic stability of the system is identified by calculating the maximum Lyapunov exponent. When a step-by-step increase in the exponent value is detected, it is determined that the filling has entered a critical rheological state and the risk level of filling migration is increased.

[0015] Furthermore, the method for constructing the mechanical response subnet includes:

[0016] A dynamic topological graph structure is constructed using shield thrust as the graph node attribute and the absolute value of the karst stratum displacement difference between adjacent monitoring points as the edge weight. The seepage correlation subnet uses a multi-scale convolution kernel to extract the frequency domain characteristics of the seawater osmotic pressure time series and predict the change trend of the cave fluid velocity.

[0017] Furthermore, the method for generating the karst pipeline rupture probability field includes:

[0018] The hydraulic connectivity strength index is mapped to the initial edge weight of the seepage association subnetwork. The transmission path of rupture risk in the karst pipeline network is simulated based on the random walk algorithm, and the spatial probability distribution field, namely the karst pipeline rupture probability field, is output.

[0019] Furthermore, the hierarchical warning triggering logic includes:

[0020] When the gradient mutation zone of the rupture probability field simultaneously meets the conditions that the continuous spatial distribution area exceeds the preset area threshold and the spatial overlap of the shield cutterhead advancement trajectory is higher than the critical value, a red warning is triggered and a grouting path topology sequence is generated. The grouting path topology sequence prioritizes the fracture channels according to the hydraulic connectivity strength index.

[0021] Furthermore, the grouting pressure distribution optimization method includes:

[0022] The rheological performance parameters of the grouting material are matched according to the risk level of filling migration. High-shear thinning slurry is used in high-risk areas, and its rheological properties meet the performance requirements of the industry standard CJJ / T 212 for grouting materials in karst formations.

[0023] Furthermore, the dynamic matching method of the rheological performance parameters includes:

[0024] A mapping relationship library between the shear dilution coefficient of grouting materials and the migration risk level of filling materials is established, and parameters are called in real time through table lookup to ensure the effective diffusion and retention of slurry in karst fissures.

[0025] Furthermore, the method further comprises:

[0026] Construct a real-time capture mechanism for cave mutation risks and a dynamic grouting preloading strategy.

[0027] The submarine karst stratum tunnel shield data monitoring and analysis system includes:

[0028] A multi-source synchronous acquisition module, which synchronously acquires shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data;

[0029] A risk three-parameter fusion module, which uses shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data based on seepage-stress coupling constraints to generate a hydraulic connectivity strength index through fracture network permeability tensor inversion. The module determines the risk level of filling migration based on the dynamic mutation characteristics of pore water pressure and integrates the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics to generate the shield disturbance entropy value.

[0030] A dual-network collaborative decision-making module, which inputs the hydraulic connectivity strength index, fill migration risk level, and shield disturbance entropy value into the spatiotemporal graph convolutional network, constructs a shield thrust-karst displacement mapping through the mechanical response subnet, predicts the cave fluid velocity through the seepage association subnet, and ultimately outputs the karst pipeline rupture probability field and shield control compensation;

[0031] Gradient mutation warning module: The gradient mutation warning module generates a graded warning signal based on the gradient mutation area distribution of the rupture probability field and its spatial position relationship with the shield cutterhead;

[0032] The grouting dynamic optimization module dynamically optimizes the grouting pressure distribution and propulsion parameters based on the shield control compensation amount and the filling migration risk level.

[0033] The technical effects and advantages of the shield data monitoring and analysis method for submarine karst stratum tunnels provided by the present invention are as follows:

[0034] The present invention generates a hydraulic connectivity index through seepage stress coupling inversion, dynamically optimizes the grouting path and rheological parameters based on the rupture probability field, realizes the coordination of shield tunneling, seepage mutation and grouting prevention and control, and fundamentally solves the problem of disconnection between prevention and control. The present invention reconstructs the hydraulic connectivity index (HCI) to accurately capture the evolution of the seepage path of hidden fissures during shield tunneling, thereby improving the identification efficiency of concealed karst pipeline water inrush channels to a new level; based on nonlinear dynamic characteristic analysis, the risk level jump signal is triggered before the irreversible migration of the filling occurs, breaking through the hysteresis shackles of traditional threshold alarms; the integration of mechanical construction entropy change, formation acoustic emission frequency variation and seawater osmotic pressure transient characteristics significantly reduces the false alarm rate of sudden surges; based on the gradient mutation characteristics of the karst pipeline rupture probability field, a graded alarm signal bound to the cutterhead spatial trajectory is generated to accurately locate the risk; according to the dynamic matching relationship between hydraulic connectivity strength and filling migration risk, the grouting rheological performance parameters and the shield advancement speed-pressure combination are autonomously optimized to ensure that the slurry achieves an intelligent phase change of "low viscosity penetration-high viscosity retention" in high-pressure fractures. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the data monitoring and analysis method for submarine karst stratum tunnel shield in Example 1;

[0036] Figure 2 This is a flow chart of the data monitoring and analysis method for submarine karst stratum tunnel shield in Example 2;

[0037] Figure 3 This is a connection diagram of the shield data monitoring and analysis system for submarine karst stratum tunnels in Example 3. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0039] Example 1:

[0040] See also Figure 1 As shown, an embodiment of the present invention provides a method for monitoring and analyzing shield data in a submarine karst stratum tunnel, the method comprising:

[0041] S1. Simultaneously collect shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data;

[0042] S2. Shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data are used to generate a hydraulic connectivity strength index through fracture network permeability tensor inversion based on seepage-stress coupling constraints. The risk level of filling migration is determined based on the dynamic mutation characteristics of pore water pressure. The shield disturbance entropy value is generated by integrating the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics.

[0043] S3. Input the hydraulic connectivity strength index, fill migration risk level, and shield disturbance entropy into the spatiotemporal graph convolutional network. A shield thrust-karst displacement mapping is constructed through the mechanical response subnet. The seepage correlation subnet predicts the karst fluid velocity. Ultimately, the karst pipeline rupture probability field and shield control compensation are output.

[0044] S4. Generate a graded warning signal based on the distribution of the gradient mutation zone of the rupture probability field and its spatial relationship with the shield cutterhead;

[0045] S5. Dynamically optimize the grouting pressure distribution and propulsion parameters based on the shield control compensation amount and the risk level of filling migration.

[0046] In step S2, the generation of the hydraulic connectivity strength index is the core of karst formation risk assessment. Specifically, this step first integrates two key data types: the three-dimensional resistivity distribution of the karst formation and the pore water pressure gradient vector field:

[0047] Three-dimensional resistivity distribution of karst formations: An acoustic emission sensor array placed around the tunnel perimeter captures changes in formation resistivity in real time, and inverts this data to form a three-dimensional resistivity cloud map (for example, areas with resistivity values ​​below 50 Ω·m are marked as water-rich dissolution zones).

[0048] Pore ​​water pressure gradient vector field: The osmotic pressure sensor group arranged at the head of the shield machine monitors the dynamics of seawater osmotic pressure and calculates the spatial distribution of the pressure gradient vector.

[0049] The shield construction machinery parameters, karst formation acoustic emission signals and seawater osmotic pressure data are fused based on the seepage-stress coupling constraint and the seepage tensor inversion algorithm. The fusion method includes:

[0050] Based on Biot's seepage-stress theory, a coupling equation is constructed to map the three-dimensional resistivity distribution into the rock porosity field, while the pore water pressure gradient field is used as the driving condition. The eigenvalues ​​of the fracture network permeability tensor are iteratively solved to quantify the degree of connectivity of the fractures in three-dimensional space. Finally, a single scalar value, the hydraulic connectivity strength index (ranging from 0 to 1), is output. The hydraulic connectivity strength index essentially represents the fluid conductivity efficiency of the karst fracture network. For example, when the index is greater than 0.7, it indicates the presence of high-speed water conduction channels between the cave groups, which may trigger the risk of piping.

[0051] The hydraulic connectivity strength index complements the subsequent analysis of pore water pressure mutation characteristics. That is, if the high connectivity index area is accompanied by a sudden pressure drop (for example, a pressure difference exceeding 0.3 MPa within 2 minutes), it will trigger a high-risk warning for filling migration, providing geomechanical boundary conditions for the calculation of shield disturbance entropy.

[0052] In the filling migration risk level determination in step S2, it is necessary to focus on analyzing the nonlinear mutation behavior of pore water pressure. The specific implementation includes:

[0053] Perform noise reduction on the pore water pressure time series data (e.g., sampling frequency 10 Hz) collected by the shield machine head sensor to extract pressure fluctuation characteristics;

[0054] The time-delay embedding method is used to reconstruct the dynamic phase space (e.g., the embedding dimension is 6 and the delay time is 3 seconds), and the one-dimensional pressure sequence is mapped into a high-dimensional system state trajectory;

[0055] Calculate the maximum Lyapunov exponent (λ) of the system. The maximum Lyapunov exponent of the system quantifies the divergence rate of adjacent tracks and is used to characterize the rheological stability of the karst filling system:

[0056] When the λ value is continuously lower than the Lyapunov exponent threshold (e.g., 0.05 bits / s), the system is in a stable state;

[0057] If a step-wise increase in λ is detected (e.g., an increase exceeding 0.35 within 30 seconds), the filling is judged to have entered a critical rheological state, at which point the filling clay particles begin to migrate under the action of the fluid.

[0058] For example, when the hydraulic connectivity strength index of a certain cave area reaches 0.75 (the high conductivity area described in Quan 2) and is accompanied by a Lyapunov exponent step (λ jumps from 0.08 to 0.52), the system raises the filling migration risk level from Level II (medium) to Level IV (high risk), indicating that large-scale filling instability and scouring may occur in this area. This filling migration risk level and the hydraulic connectivity strength index jointly constitute the calculation constraint of the shield disturbance entropy value (for example, the cutterhead speed fluctuation needs to be suppressed in high-risk areas).

[0059] It should be noted that the risk level of filling migration is set artificially according to the situation, such as level I-IV or level I-V.

[0060] The core of the spatiotemporal graph convolutional network in step S3 lies in the dual-proton network collaboration; the dual-proton network includes a mechanical response subnet and a percolation-related subnet;

[0061] The mechanical response subnet construction method includes:

[0062] The monitoring points of each ring propulsion unit of the shield machine are used as graph nodes (for example, one displacement monitoring section is set every five rings), and the node attributes are assigned the real-time shield propulsion force (unit: kN);

[0063] The absolute difference of the displacement monitoring values ​​of the karst strata between adjacent nodes is used as the edge weight (for example, when the displacement difference between sections A and B reaches 0.5 mm, the weight is set to 0.8) to characterize the coordination of stratum deformation;

[0064] By learning the nonlinear relationship between thrust and displacement through graph convolution (for example, an 800kN thrust induces a sudden increase in displacement at the cave boundary), the shield thrust-karst displacement mapping function is output to predict the rock deformation in key areas.

[0065] The operation methods of the percolation association subnetwork include:

[0066] Perform wavelet transformation on the seawater osmotic pressure time series collected by S1 (e.g., using a 0.1-5 Hz bandpass filter) and decompose the frequency domain components using convolution kernels of different scales (e.g., long-term kernels to extract tidal cycle characteristics, short-term kernels to capture pressure pulses);

[0067] By integrating the hydraulic connectivity strength index with frequency domain characteristics, the changing trend of fluid velocity in the cave is predicted (for example, when the amplitude of the low-frequency component increases by 3 times and the high-frequency pulses are dense, it indicates that the flow velocity will exceed 2m / s).

[0068] Finally, the dual-proton network jointly outputs the karst pipeline rupture probability field (presented as a 0-1 probability cloud map) and the shield control compensation amount (such as reducing the cutterhead speed by 5% and increasing the grouting pressure by 0.2MPa). For example, when the displacement mapping value of a certain cave area exceeds the warning threshold and the predicted fluid flow rate is greater than 1.5m / s, the rupture probability of this location rises to 0.92.

[0069] In the optimization design of the seepage-related subnetwork, the spatial transmission mechanism of the hydraulic connectivity intensity index is introduced. The spatial transmission mechanism includes:

[0070] The generated hydraulic connectivity strength index (0-1 scalar) is mapped to the initial edge weight of the seepage subnetwork to construct a spatial conduction model of the karst conduit network (for example, the edge weight between caves with an index > 0.7 is set to 0.9, indicating strong hydraulic connection);

[0071] Simulate the diffusion of rupture risk in a pipeline network based on a random walk algorithm:

[0072] Taking the shield tunneling face as the starting point, the risk is transferred along the edge weight gradient direction (for example, an edge with a weight of 0.9 has a 90% probability of becoming a transmission path);

[0073] When the path passes through the high displacement deformation zone (output of the mechanical response subnet), the rupture risk value is exponentially amplified according to the deformation amount;

[0074] After millions of walk iterations, the spatial probability distribution field of karst pipeline rupture (three-dimensional probability cloud map) is output.

[0075] This spatial transmission mechanism significantly improves the ability to capture hidden geological risks. For example, when the random walk path forms a closed-loop transmission in area A (edge ​​weights are all > 0.8), even if the initial seepage pressure in this area is stable, the system will still mark it as a Level IV risk area, triggering an adaptive 10% reduction in the shield cutter head torque.

[0076] The hierarchical warning mechanism of step S4 is triggered by dual spatial coupling analysis, including:

[0077] Based on the spatial probability field of karst pipeline rupture, the probability gradient modulus (the rate of change of probability per unit distance) is calculated, and continuous areas where the gradient value exceeds the critical threshold (such as the area with gradient > 0.35 / m), namely the gradient mutation area, are extracted;

[0078] When the continuous distribution area of ​​the gradient mutation zone exceeds the safety tolerance (for example, >15 m2, indicating the risk of large-scale structural instability), the spatial overlap between the mutation zone and the shield cutterhead advancement trajectory is calculated in real time (for example, the range of 3 times the diameter in front of the cutterhead). When the overlap rate exceeds the critical value (for example, >60%), direct risk intervention is determined; direct risk intervention includes red warning response;

[0079] Red Alert Response:

[0080] The hydraulic connectivity strength index is used as the priority ranking basis. The higher the index, the higher the channel ranking (for example, fracture channels with an index of 0.85 are given priority for grouting).

[0081] Perform high-pressure compensatory grouting on the first N channels in the sequence (the pressure increase is positively correlated with the probability of rupture).

[0082] For example, when a tunnel was excavated to loop number K102+360, a sudden gradient change (peak gradient 0.48 / m) covering 22 square meters was detected on the right side of a cluster of karst caves. This area overlapped 73% with the cutterhead trajectory. The system triggered a red alert and generated a grouting sequence: prioritizing primary fractures with a hydraulic index of 0.92 (increasing grouting pressure to 0.8 MPa), followed by secondary channels with an index of 0.78. After this implementation, the displacement increment in this area decreased by 82%, verifying the effectiveness of the path sequencing.

[0083] In the grouting execution phase of step S5, a dynamic adaptation mechanism for rheological properties is implemented based on the generated grouting path topology sequence. The dynamic adaptation mechanism for rheological properties includes risk-material mapping rules and performance parameter adaptation:

[0084] Risk-Material Mapping Rules:

[0085] Match grouting materials with different rheological properties based on the filling migration risk level (e.g., level I-IV) divided by the fracture probability field;

[0086] In high-risk areas (probability > 0.8), artificially set, for example, level III-IV high risk, the corresponding fracture probability field area is a high-risk area, and high shear dilution slurry is used. Its characteristics meet the requirements of the "Code for Geotechnical Engineering Investigation of Urban Rail Transit" CJJ / T 212 for karst strata, including:

[0087] Low shear rate is Newtonian fluid (easy to penetrate deep into cracks);

[0088] A sharp drop in viscosity at high shear rates (ensuring minimal drag during high-pressure grouting);

[0089] The structural viscosity is quickly restored after standing (to prevent the slurry from being washed away by groundwater).

[0090] Adaptive performance parameters: The rheological parameters (viscosity, yield stress) of the slurry are adjusted according to the risk level. For example, a customized slurry is used in a level IV risk area (probability of rupture 0.92), including:

[0091] Initial viscosity ≤50 mPa·s (to ensure permeability); shear rate >100 s -1 The viscosity drops to less than 30% of the initial value (adapting to grouting pressure of more than 0.8MPa); the water separation rate is less than 3% (in line with the durability requirements of CJJ / T 212).

[0092] Engineering verification: In the high-risk area K102+360 (original displacement increment 1.2 mm / d), after injecting high-shear dilution slurry, the slurry diffusion radius reached 8.5 m when the grouting pressure reached 0.8 MPa (conventional slurry is only 4 m). After 3 days, the displacement increment stabilized at 0.2 mm / d, and core sampling of the slurry showed that it was tightly bonded to the cave wall.

[0093] Based on the adaptation of rheological properties, a dynamic mapping mechanism of shear dilution coefficient is constructed to achieve precise control. The control methods include:

[0094] The shear dilution factor is defined as the ratio of the viscosity of the slurry at high or low shear rates (e.g. 100s -1 with 1s -1 Viscosity ratio), establish a strong correlation between the shear dilution coefficient and the filling migration risk level (taking levels I-IV as an example). For example, in high-risk areas (Level IV), the shear dilution coefficient should be greater than 3.0 (the viscosity drops sharply under high shear, which is conducive to high-pressure injection), and in low-risk areas (Level I), the shear dilution coefficient should be maintained at 1.2-1.8 (to maintain moderate fluidity), thus forming a "Karst Risk-Rheological Parameter Mapping Relationship Library";

[0095] Exemplary:

[0096] When a red alert is triggered, the system automatically searches a database based on the target fracture's filling migration risk level. It then calls upon the corresponding rheological parameters to drive the grouting equipment (for example, a slurry formula with a coefficient of 3.2 and a viscosity of 45 mPa·s for a Class IV region is automatically matched). The grouting pressure and advancement speed are then adjusted simultaneously to ensure that the slurry completes the "infiltration-retention" transition within the fracture.

[0097] Example 2:

[0098] like Figure 2 As shown, this embodiment further improves the design based on the first embodiment. The difference is that in the actual operation of the first embodiment, it was found that the random walk algorithm did not respond adequately to the sudden seepage changes in isolated, unnetworked caves, and the mechanical response subnet had a prediction delay (about 2-3 minutes) for the instantaneous displacement increment induced by the sudden change in pore water pressure. This resulted in delayed grouting intervention in high-risk areas and failed to effectively suppress the chain reaction of small-scale piping. Based on this, the data monitoring and analysis method for submarine karst stratum tunnel shield also includes:

[0099] S6. Construct a real-time capture mechanism for cave mutation risk and a dynamic preloading strategy for grouting.

[0100] The real-time capture mechanism of cave mutation risk includes:

[0101] An active acoustic wave transmitting array is deployed in a 120° fan-shaped area in front of the shield cutterhead, emitting sweep frequency pulses every 10 seconds;

[0102] Identify isolated caves that are not connected to the network by the attenuation slope of the reflected signal (if the attenuation rate is greater than 30dB / m, the area is marked as a potential isolated cave);

[0103] When the standard deviation of pore water pressure fluctuations in an isolated cave suddenly increases (e.g., a 200% increase within 30 seconds) and is accompanied by a Lyapunov exponent λ > 0.4, a purple warning for isolated cave piping is triggered (a new independent warning level).

[0104] Establish pore water pressure variation and karst displacement increment The time-delay transfer function includes:

[0105] ;

[0106] Where, is the time lag constant, which is the field calibration value (e.g. τ = 110 seconds for seabed clay). For the current moment, is the displacement response coefficient, which represents the sensitivity of the formation to the seepage pressure. is the displacement attenuation factor, is the time integral variable, and e is a natural constant.

[0107] When real-time monitoring When the pressure threshold is exceeded (e.g., pressure drop of 0.25 MPa in 2 minutes), the displacement increment after 3 minutes is predicted based on the transfer function. >0.8mm, the grouting system will be activated in advance.

[0108] Grouting dynamic preloading strategies include:

[0109] When the purple warning is triggered, immediately inject high thixotropic plugging slurry (for example, static shear force ≥ 15Pa, dynamic plasticity ratio > 1.2) into the isolated hole.

[0110] Calculate the pre-grouting amount based on 50% of the predicted displacement increment (e.g. =1.0mm, inject 0.5m 3 slurry);

[0111] When the hydraulic connectivity strength index is greater than 0.6, the pulse grouting mode (0.5 Hz, peak pressure = 1.2 × conventional value) is adopted.

[0112] Example 3:

[0113] like Figure 3 As shown, based on the same inventive concept as the submarine karst formation tunnel shield data monitoring and analysis method in the aforementioned embodiment, the present application provides a submarine karst formation tunnel shield data monitoring and analysis system. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0114] Multi-source synchronous acquisition module, which synchronously collects shield construction machinery parameters, karst formation acoustic emission signals and seawater osmotic pressure data;

[0115] The risk three-parameter fusion module combines shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data based on seepage-stress coupling constraints. It generates a hydraulic connectivity strength index through fracture network permeability tensor inversion. The risk level of filling migration is determined based on the dynamic mutation characteristics of pore water pressure. The cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics are integrated to generate the shield disturbance entropy value.

[0116] The dual-network collaborative decision-making module inputs the hydraulic connectivity strength index, fill migration risk level, and shield disturbance entropy value into the spatiotemporal graph convolutional network. It constructs a shield thrust-karst displacement mapping through the mechanical response subnet and predicts the cave fluid velocity through the seepage association subnet. It ultimately outputs the karst pipeline rupture probability field and shield control compensation.

[0117] Gradient mutation warning module: The gradient mutation warning module generates a graded warning signal based on the gradient mutation area distribution of the rupture probability field and its spatial position relationship with the shield cutterhead;

[0118] Grouting dynamic optimization module, which dynamically optimizes grouting pressure distribution and propulsion parameters based on shield control compensation and filling migration risk level.

[0119] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0120] The above is only a preferred specific implementation method of the embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes based on the technical solution and concept of the present application within the technical scope disclosed in the present application, and they should be covered by the scope of protection of the present application.

Claims

1. A method for monitoring and analyzing shield tunneling data in submarine karst strata, characterized in that: include: Simultaneously collect shield construction machinery parameters, karst formation acoustic emission signals and seawater osmotic pressure data; Based on the seepage-stress coupling constraint, the parameters of shield construction machinery, karst formation acoustic emission signals, and seawater osmotic pressure data were used to generate a hydraulic connectivity strength index through fracture network permeability tensor inversion. The risk level of filling migration was determined based on the dynamic mutation characteristics of pore water pressure, and the shield disturbance entropy value was generated by integrating the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics. The hydraulic connectivity strength index, filling material migration risk level and shield disturbance entropy value are input into the spatiotemporal graph convolutional network, and the shield thrust-karst displacement mapping is constructed through the mechanical response subnet. The seepage association subnet is used to predict the cave fluid velocity, and finally the karst pipeline rupture probability field and the shield control compensation are output. The method for generating the karst pipeline rupture probability field includes: mapping the hydraulic connectivity strength index into the initial edge weight of the seepage association subnet, simulating the transmission path of the rupture risk in the karst pipeline network based on the random walk algorithm, and outputting the spatial probability distribution field, namely the karst pipeline rupture probability field. The method for constructing the mechanical response subnet includes: using the shield thrust as the graph node attribute and the absolute value of the karst stratum displacement difference between adjacent monitoring points as the edge weight to construct a dynamic topological graph structure. The seepage association subnet uses a multi-scale convolution kernel to extract the frequency domain characteristics of the seawater osmotic pressure time series to predict the trend of cave fluid velocity changes. According to the distribution of gradient mutation zones of the rupture probability field and their spatial relationship with the shield cutterhead, a graded warning signal is generated. Based on the shield control compensation amount and the filling migration risk level, the grouting pressure distribution and advancement parameters are dynamically optimized. The grouting pressure distribution optimization method includes: matching the rheological performance parameters of the grouting material according to the filling migration risk level. High-shear thinning slurry is used in high-risk areas. Its rheological properties meet the performance requirements of the industry standard CJJ / T 212 for grouting materials in karst formations.

2. The method for monitoring and analyzing shield tunneling data in submarine karst strata according to claim 1, characterized in that: The reconstruction method of hydraulic connectivity strength index includes: The three-dimensional resistivity distribution of the karst formation is coupled with the pore water pressure gradient vector field, and the hydraulic connectivity strength index is generated using the permeability tensor inversion algorithm. The permeability tensor inversion algorithm is constructed based on Biot's seepage-stress theory, and its output value is used to characterize the fluid conductivity efficiency of the karst fracture network.

3. The method for monitoring and analyzing shield tunneling data in submarine karst strata according to claim 1, characterized in that: The method for determining the risk level of filling migration includes: The nonlinear dynamic phase space of the pore water pressure time series is reconstructed, and the sudden change in the dynamic stability of the system is identified by calculating the maximum Lyapunov exponent. When a step-by-step increase in the exponent value is detected, it is determined that the filling has entered a critical rheological state and the risk level of filling migration is increased.

4. The method for monitoring and analyzing shield tunneling data in submarine karst strata according to claim 1, characterized in that: The hierarchical warning trigger logic includes: When the gradient mutation zone of the rupture probability field simultaneously meets the conditions that the continuous spatial distribution area exceeds the preset area threshold and the spatial overlap of the shield cutterhead advancement trajectory is higher than the critical value, a red warning is triggered and a grouting path topology sequence is generated. The grouting path topology sequence prioritizes the fracture channels according to the hydraulic connectivity strength index.

5. The method for monitoring and analyzing shield tunneling data in submarine karst strata according to claim 1, characterized in that: The dynamic matching method of the rheological performance parameters includes: A mapping relationship library between the shear dilution coefficient of grouting materials and the migration risk level of filling materials is established, and parameters are called in real time through table lookup to ensure the effective diffusion and retention of slurry in karst fissures.

6. The method for monitoring and analyzing shield tunneling data in submarine karst strata according to claim 1, characterized in that: The method also includes: Construct a real-time capture mechanism for cave mutation risks and a dynamic grouting preloading strategy.

7. A submarine karst stratum tunnel shield data monitoring and analysis system, which implements the submarine karst stratum tunnel shield data monitoring and analysis method according to any one of claims 1 to 6, characterized in that the system include: A multi-source synchronous acquisition module, which synchronously acquires shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data; A risk three-parameter fusion module, which uses shield construction machinery parameters, karst formation acoustic emission signals, and seawater osmotic pressure data based on seepage-stress coupling constraints to generate a hydraulic connectivity strength index through fracture network permeability tensor inversion. The module determines the risk level of filling migration based on the dynamic mutation characteristics of pore water pressure and integrates the cutterhead torque spectrum characteristics and propulsion force fluctuation characteristics to generate the shield disturbance entropy value. A dual-network collaborative decision-making module, which inputs the hydraulic connectivity strength index, the filling migration risk level, and the shield disturbance entropy value into the spatiotemporal graph convolutional network, constructs a shield thrust-karst displacement mapping through the mechanical response subnet, predicts the cave fluid velocity through the seepage association subnet, and finally outputs the karst pipeline rupture probability field and the shield control compensation amount; the method for generating the karst pipeline rupture probability field includes: mapping the hydraulic connectivity strength index into the initial edge weight of the seepage association subnet, simulating the transmission path of the rupture risk in the karst pipeline network based on the random walk algorithm, and outputting a spatial probability distribution field, namely the karst pipeline rupture probability field; the method for constructing the mechanical response subnet includes: using the shield thrust as the graph node attribute and the absolute value of the karst stratum displacement difference between adjacent monitoring points as the edge weight to construct a dynamic topological graph structure; the seepage association subnet uses a multi-scale convolution kernel to extract the frequency domain characteristics of the seawater osmotic pressure time series to predict the trend of cave fluid velocity changes; Gradient mutation warning module: The gradient mutation warning module generates a graded warning signal based on the gradient mutation area distribution of the rupture probability field and its spatial position relationship with the shield cutterhead; A dynamic grouting optimization module dynamically optimizes grouting pressure distribution and propulsion parameters based on shield control compensation and filling migration risk level. The grouting pressure distribution optimization method includes: matching the rheological performance parameters of the grouting material according to the filling migration risk level, using a high shear-thinning slurry in high-risk areas, and its rheological properties meet the performance requirements of the industry standard CJJ / T 212 for grouting materials in karst formations.

Images