Seismic electric coupling advanced detection TBM tunneling parameter intelligent self-tuning system

Abstract

This invention discloses an intelligent self-tuning system for TBM tunneling parameters based on seismic-electric coupling advanced detection, belonging to the technical field of tunnel boring equipment. The system constructs a closed-loop control architecture through the collaborative work of five modules: "seismic-electric advanced detection → multi-source data processing → geological condition identification → tunneling parameter tuning → feedback model update." The seismic-electric coupling advanced detection module synchronously acquires seismic wave and electrical response data. After correction and noise reduction by the multi-source data synchronous acquisition and preprocessing module, the geological condition fusion identification and risk assessment module completes the surrounding rock condition determination and risk assessment. The tunneling parameter self-tuning control module adaptively adjusts the tunneling parameters based on the risk results. This invention achieves intelligent linkage between geological detection and tunneling control, improving the safety, adaptability, and tunneling efficiency of the TBM tunneling process, and solving the technical defects of traditional TBMs, such as single geological detection, lagging parameter adjustment, and lack of closed-loop optimization.

Description

Technical Field

[0001] This invention relates to the field of tunnel boring equipment technology, specifically to an intelligent self-tuning system for advanced detection of TBM tunneling parameters using seismic-electric coupling. Background Technology

[0002] As the core equipment in tunnel construction, the tunneling efficiency and construction safety of TBMs directly depend on the accuracy of the prediction of the geological conditions ahead and the adaptability of the tunneling parameters. Traditional TBMs often use single technologies for advanced geological exploration (such as seismic wave detection and electrical resistivity tomography), which suffer from problems such as limited data dimensions and insufficient reliability of detection results: seismic wave detection is sensitive to the integrity of the surrounding rock structure but is greatly affected by mechanical vibration; electrical resistivity tomography shows obvious results in terms of water content but is easily affected by electromagnetic coupling interference, and the two types of data often cannot be synchronized in space and time, resulting in large errors in geological condition identification.

[0003] Meanwhile, traditional TBMs rely heavily on operator experience or simple preset logic for adjusting tunneling parameters, lacking real-time linkage with geological survey results. When encountering complex geological conditions (such as fractured zones, water-rich areas, and abrupt lithological changes), parameter adjustments lag behind, easily leading to safety risks such as cutterhead jamming, excessive cutter wear, and tunnel collapse, or resulting in low tunneling efficiency due to conservative parameters. Furthermore, existing systems lack effective feedback optimization mechanisms, failing to dynamically correct the detection model and parameter tuning logic based on actual tunneling results, making it difficult to continuously improve accuracy after long-term use.

[0004] Therefore, there is an urgent need for a TBM system that can achieve simultaneous acquisition of multi-source geological data, accurate identification of geological conditions, intelligent adaptation of tunneling parameters, and continuous optimization, in order to solve the technical defects of existing technologies such as single detection, lagging parameter adjustment, and lack of closed-loop optimization. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent self-tuning system for TBM tunneling parameters using seismoelectric coupling for advanced detection, in order to solve the problems of current advanced geological detection methods being limited, tunneling parameter adjustments being disconnected from geological conditions, and the lack of a continuous adaptive tuner.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: The seismic-electric coupling advanced detection module is installed on the cutterhead, front shield, or circumferential direction of the TBM. It adopts an integrated design of "seismic wave detection + electrical detection" and includes a seismic wave detection unit and an electrical detection unit. The seismic wave detection unit consists of an active excitation device and a multi-channel vibration sensor array. The active excitation device generates seismic waves of a specific frequency, and the multi-channel vibration sensor array collects the seismic wave response data of the surrounding rock in front. The electrical detection unit consists of a current injection electrode and a response electrode. The current injection electrode injects a stable current into the surrounding rock, and the response electrode collects electrical response data of the surrounding rock (such as resistivity, polarizability, etc.). A unified detection step distance (e.g., 0.5-2m / step) and a synchronous triggering mechanism are adopted. The TBM main control system sends a synchronous triggering signal to ensure that seismic wave excitation, vibration response acquisition, current injection, and electrical response acquisition are completed within the same tunneling cycle. This ensures that the seismic wave response data and the electrical response data correspond one-to-one in terms of spatial location (same detection section) and time scale (same tunneling cycle).

[0007] The multi-source data synchronous acquisition and preprocessing module is connected to the seismic-electric coupling advanced detection module via industrial Ethernet or CAN bus. Its core functions include: Time synchronization alignment: GPS time synchronization or local high-precision clock synchronization technology is used to timestamp and calibrate the seismic response data and electrical response data to ensure the time consistency of the two types of data; Joint correction: Collect real-time tunneling parameters of the TBM (thrust, cutterhead speed, torque, etc.), construct a mapping relationship model between tunneling parameters and seismic-electric response signals (such as modeling methods based on BP neural networks and support vector machines), separate the mechanical vibration interference component in the seismic response data and the electromagnetic coupling interference component in the electrical response data through this model, and compensate and correct the effective signal after interference, finally generating high-quality seismic-electric coupling feature data.

[0008] The geological condition fusion identification and risk assessment module is based on machine learning algorithms (such as random forests and deep learning models) to build a fusion identification model. The specific process is as follows: Feature extraction: Key parameters are extracted from the seismic-electric coupling feature data. Seismic wave feature parameters include wave velocity, amplitude attenuation coefficient, and dominant frequency offset, while electrical feature parameters include mean resistivity, resistivity variation coefficient, and polarizability. Joint Judgment: The surrounding rock condition is identified based on the mutual corroboration relationship between seismic wave characteristic parameters and electrical characteristic parameters. If the two types of parameters show a consistent trend (e.g., wave velocity decreases while resistivity decreases, corresponding to a fractured and water-rich area), the risk confidence of the geological condition is increased (e.g., confidence increases from 0.7 to 0.9). If the two types of parameters show contradictory trends (e.g., wave velocity decreases but resistivity increases), the risk confidence of the corresponding geological condition is decreased (e.g., from 0.7 to 0.3) or it is marked as an abnormal geological condition, triggering a manual review prompt. Output results: The final output is the geological state type of the surrounding rock (such as intact hard surrounding rock, fractured surrounding rock, water-rich surrounding rock, lithological abrupt change section, etc.) and the corresponding risk confidence level (0-1.0, where 0.8-1.0 is extremely high risk, 0.6-0.8 is high risk, 0.4-0.6 is medium risk, and 0-0.4 is low risk).

[0009] The tunneling parameter self-tuning control module is connected to the TBM's hydraulic control system and variable frequency drive system to achieve adaptive control of tunneling parameters. Proportional control logic: Establish a continuous functional relationship between the adjustment range of tunneling parameters and the risk confidence level. For example, the adjustment range of thrust ΔF = k1 × (C - 0.4) (where k1 is the proportional coefficient and C is the risk confidence level), and the adjustment range of cutterhead speed Δn = k2 × (0.4 - C) (where k2 is the proportional coefficient). This makes the parameter adjustment change continuously with the risk confidence level, avoiding equipment shock caused by discrete adjustment. Safety control strategy: When a high-risk (0.6-0.8) or extremely high-risk (0.8-1.0) geological condition is identified, safety control is automatically implemented: the single-step advance is reduced from the usual 1.5-2m to 0.5-1m, the upper limit of torque is limited to 60%-80% of the rated torque, and if the risk confidence level continues to be higher than 0.9, a shutdown warning is triggered, and manual investigation is required.

[0010] The tunneling feedback and model self-updating module collects actual tunneling response data through sensors, including penetration changes, torque fluctuations, tool wear, and propulsion resistance, to achieve dynamic model updates. Fusion identification model update: Calculate the deviation between the actual tunneling response data and the geological condition fusion identification results (e.g., the predicted intact surrounding rock but the actual penetration fluctuates greatly), and dynamically adjust the weight parameters of seismic-electric coupling features in the fusion identification model based on the deviation value (e.g., increase the weight of electrical feature parameters to improve the accuracy of water content identification). Tuning model update: Based on the changes in penetration depth, torque fluctuations, and tool wear information, the proportional coefficient of the tunneling parameter tuning model is corrected (e.g., when the tool wear exceeds the threshold, the propulsion proportional coefficient k1 is reduced to avoid excessive wear), ensuring the adaptability of subsequent tunneling parameters.

[0011] Through the sequential linkage and reverse feedback of the above five modules, a closed-loop TBM tunneling control system of "detection-processing-identification-tuning-updating" is formed, realizing real-time dynamic matching of geological conditions and tunneling parameters.

[0012] The present invention has the following beneficial effects: This system effectively compensates for the limitations of a single detection method and improves the integrity and reliability of the surrounding rock data by combining seismic-electric coupling dual-source synchronous detection with joint correction of tunneling working parameters. Based on the mutual verification logic of seismic waves and electrical characteristics, the identification of geological conditions and the assessment of risk confidence are made more scientific. Through the continuous proportional tuning of tunneling parameters driven by risk confidence and the safety control strategy under high-risk conditions, the dynamic and precise adaptation of tunneling parameters and geological conditions is achieved, reducing equipment impact and construction safety hazards. By combining a closed-loop feedback model self-updating mechanism, the detection, identification, and tuning models can be continuously optimized based on the actual tunneling response, thereby continuously improving the system's adaptability to complex geological conditions and comprehensively ensuring the safety, stability, and efficiency of the TBM tunneling process. Attached Figure Description

[0013] Figure 1 This is a system block diagram of the present invention; Figure 2 The diagram illustrates the specific steps of each module in the system of this invention. Detailed Implementation

[0014] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention. (I) System Hardware Configuration

[0015] Vibration-electric coupling advanced detection module: Four sets of detection units are evenly arranged around the edge of the TBM cutterhead. Each set includes one hydraulic active excitation device (excitation frequency 50-500Hz), two high-sensitivity vibration sensors (range 0-100g, frequency response 10-1000Hz), two injection electrodes (copper alloy material, rated current 10A), and two response electrodes (accuracy ±0.1%). The detection step distance is set to 1m / step. The synchronous trigger signal is issued by the TBM main control system (PLC model S7-1500), and the trigger interval is consistent with the tunneling cycle (10-20s / cycle).

[0016] Multi-source data synchronous acquisition and preprocessing module: A data acquisition card (sampling rate 1MHz, resolution 16-bit) is used to synchronously acquire seismic and electrical data, and time alignment is achieved through a GPS timing module (time synchronization accuracy ±1μs); tunneling working condition parameters are acquired through the TBM's existing sensors (propulsion force sensor range 0-5000kN, cutterhead speed sensor accuracy ±0.1r / min, torque sensor range 0-1000kN•m); the mapping relationship model adopts a BP neural network (input layer is working condition parameters, hidden layer has 10 neurons, output layer is interference compensation), and data processing is implemented through Python programming.

[0017] Geological State Fusion Identification and Risk Assessment Module: The fusion identification model is run on an industrial computer (CPU i7-12700, memory 32GB). The model is based on a deep learning network built on the TensorFlow framework. The input layer consists of 8-dimensional feature parameters (3 seismic wave parameters and 5 electrical parameters), and the output layer consists of 6 types of geological states (intact and hard surrounding rock, relatively intact surrounding rock, fractured surrounding rock, water-rich fractured surrounding rock, lithological abrupt change section, and abnormal geological section) and risk confidence. The model training samples come from tunnel construction data of 3 different geological conditions (sample size 10,000 sets).

[0018] Tunneling parameter self-tuning control module: communicates with the TBM hydraulic control system and variable frequency drive system via Profinet bus, with proportional coefficient k1 set to 500kN and k2 set to 5r / min; in the safety control strategy, the single-step advance is set to 0.8m under high-risk conditions, the torque limit is set to 70% of the rated torque, and the PLC emergency stop warning is triggered under extremely high-risk conditions.

[0019] Tunneling feedback and model self-updating module: Three new tool wear sensors (accuracy ±0.1mm) are installed on the cutterhead tool holder. The penetration depth is calculated by the propulsion displacement sensor (range 0-2m, accuracy ±0.1mm) and time. The model update cycle is set to every 10 tunneling cycles (approximately 100-200s). The weight adjustment adopts the gradient descent method with a correction step size of 0.01. (II) Core Model Work Management

[0020] 1. Working principle of the seismic-electric coupling advanced detection module This module is based on the dual-physics field detection principle of "seismic wave propagation characteristics + electrical response laws": Seismic wave detection principle: The active excitation device emits elastic seismic waves of a specific frequency into the surrounding rock. When the seismic waves propagate in surrounding rocks with different integrity and lithology, parameters such as wave velocity, amplitude attenuation coefficient, and dominant frequency will change regularly (e.g., wave velocity decreases and amplitude attenuation accelerates in fractured surrounding rock). The multi-channel vibration sensor array captures these changes and converts them into seismic wave response data, enabling a preliminary judgment on the structural integrity and lithology of the surrounding rock. Electrical detection principle: A stable DC or low-frequency AC current is injected into the surrounding rock by the injection electrode. The conduction characteristics of the current in the surrounding rock are closely related to the water content and porosity of the surrounding rock (e.g., the resistivity of water-rich surrounding rock is significantly lower than that of dry surrounding rock). The response electrode collects the potential difference at both ends of the surrounding rock and calculates electrical parameters such as resistivity and polarizability to achieve accurate perception of the water content. Synchronous triggering principle: Due to the significant difference between the propagation speed of seismic waves (km / s) and the conduction speed of current (near the speed of light), if the triggering is not synchronized, the two types of data in the same detection section will be misaligned in time. Therefore, a unified triggering signal is issued by the main control system to ensure that the seismic excitation and current injection start synchronously, and the vibration acquisition and electrical acquisition end synchronously. This ultimately achieves a one-to-one correspondence of data in space (same detection step section) and time (same tunneling cycle), providing a common source data foundation for subsequent fusion identification.

[0021] 2. Working principle of multi-source data synchronous acquisition and preprocessing module The core principle is based on the correlation between operating conditions and interference and the principle of data completion. Time synchronization principle: The GPS timing module is used to realize cross-module timestamp calibration, so that the time error between the seismic response data and the electrical response data is controlled within the microsecond level, avoiding feature matching errors caused by time misalignment; Interference Separation and Compensation Principle: During TBM tunneling, changes in thrust and cutterhead speed generate mechanical vibrations, and the operation of the equipment motor produces electromagnetic interference. These interferences are superimposed on the original detection data. By constructing a mapping model between tunneling operating parameters and interference signals, and using a BP neural network to learn the correlation between operating parameters (such as the synchronous increase in mechanical vibration interference intensity when thrust increases) and interference signals, the interference components in the original data can be accurately predicted. Then, through the calculation logic of "original data - interference prediction value + effective signal compensation value", the interference is stripped away and the effective signal is supplemented, ultimately generating seismic-electric coupling characteristic data that truly reflects the characteristics of the surrounding rock.

[0022] 3. Working principle of the geological condition fusion identification and risk assessment module Based on the principle of "feature mutual verification + confidence quantification": The principle of mutual verification of characteristics: Seismic wave characteristic parameters (wave velocity, amplitude attenuation, etc.) mainly reflect the physical and mechanical properties of the surrounding rock (hardness, integrity), while electrical characteristic parameters (resistivity, polarizability, etc.) mainly reflect the hydrological and porosity characteristics of the surrounding rock (water content, porosity). The physical meanings of the two types of parameters are complementary. When the state of the surrounding rock changes (such as entering a water-rich fractured zone), it will simultaneously lead to a decrease in seismic wave velocity and a decrease in resistivity. The two types of characteristics change in a consistent manner, indicating that the geological state determination has dual physical basis and is more reliable. If a contradictory situation occurs where the seismic wave velocity decreases (indicating fracture) but the resistivity increases (indicating dryness), it indicates the possible existence of unknown geological factors (such as the presence of gypsum interlayers), and the reliability of the determination needs to be reduced. Risk confidence quantification principle: The model is trained on historical data through deep learning. The matching thresholds of two types of features under different geological conditions are preset. When the features are consistent, the confidence is increased by superimposing the preset weights. When the features are contradictory, the confidence is decreased or marked as abnormal according to the degree of deviation. Finally, the confidence value in the range of 0-1 is output to quantify the reliability of the geological condition judgment.

[0023] 4. Working principle of the tunneling parameter self-tuning control module Based on the principles of "risk-parameter adaptation" and "stable control": Proportional control principle: The numerical change of risk confidence reflects the gradual process of geological conditions (such as the transition from low risk to high risk). Therefore, the adjustment of tunneling parameters must follow the principle of "gradual adaptation". By establishing a continuous functional relationship between the adjustment range and the risk confidence, parameters such as thrust and cutterhead rotation speed can change smoothly with the risk level, avoiding cutterhead impact and increased surrounding rock disturbance caused by sudden parameter changes. Safety control principle: In high-risk or extremely high-risk geological conditions (such as water-rich fractured zones), the surrounding rock has low bearing capacity and poor stability. In this case, it is necessary to reduce the disturbance range of the surrounding rock by "reducing the step distance" and "limiting torque" to avoid equipment jamming or tool breakage caused by cutterhead overload. When the risk exceeds the safety threshold, a shutdown warning is triggered. In essence, it is based on the adaptation logic of "risk level - safety redundancy" to ensure construction safety.

[0024] 5. Working principle of the tunneling feedback and model self-updating module Based on the principle of "closed-loop iterative optimization": The principle of updating the fusion identification model: The actual tunneling response data (such as penetration fluctuations and torque changes) is a direct reflection of the true state of the surrounding rock. If the actual penetration fluctuations far exceed the range predicted by the fusion identification results (such as the prediction of intact surrounding rock but the actual fluctuations are severe), it indicates that the weight settings of the seismic-electric coupling features are unreasonable (such as not fully considering the sensitivity of electrical features to water content). By calculating the deviation value, the feature weights are dynamically adjusted using the gradient descent method to make the subsequent identification results of the model closer to reality. The principle of the tuning model update: The proportional coefficient of the tunneling parameter tuning model is preset based on conventional geological conditions. When there is increased tool wear or abnormal torque fluctuation, it indicates that the parameter adjustment range does not match the actual working conditions (such as excessive propulsion force leading to excessive tool wear). Therefore, based on feedback data such as penetration depth and tool wear, the proportional coefficient is corrected so that subsequent parameter adjustments can ensure tunneling efficiency and reduce equipment wear, achieving closed-loop optimization of "prediction-execution-feedback-correction". (III) System Workflow

[0025] Exploration Phase: During the TBM tunneling process, the main control system sends a synchronous trigger signal every 10 seconds, the active excitation device excites 50-500Hz vibration waves, and the vibration sensor collects the vibration wave response data; at the same time, the injection electrode injects a stable current of 10A, and the response electrode collects the electrical response data. Both types of data are transmitted to the data acquisition module in real time.

[0026] Data processing stage: The data acquisition module completes time alignment through GPS timing, combines the operating parameters such as propulsion force and cutter head speed collected at the same time, and inputs them into the BP neural network model to separate mechanical vibration and electromagnetic coupling interference, and outputs vibration-electric coupling characteristic data (such as wave velocity 3500m / s, resistivity 1000Ω•m).

[0027] Geological identification stage: The fusion identification model receives feature data, extracts 8 key parameters such as wave velocity and resistivity, and judges whether the changing trends of the two types of features are consistent; if consistent (e.g., high wave velocity, high resistivity), it outputs "intact and hard surrounding rock" and a risk confidence level of 0.3 (low risk); if contradictory, it lowers the confidence level or marks it as an anomaly.

[0028] Parameter tuning stage: The tuning module calculates the adjustment range based on the risk confidence level. When the risk is low (0.3), the propulsion force is adjusted to 2950kN and the cutter head speed is adjusted to 6.5r / min, maintaining a single-step advance of 1.5m. If the risk is high (0.75), the single-step advance is automatically reduced to 0.8m and the torque limit is set at 700kN•m. If the risk exceeds 0.9, a shutdown warning is triggered.

[0029] Feedback Update Phase: The feedback module collects actual data such as penetration depth, torque fluctuation, and tool wear, and calculates the deviation from the recognition result. If the deviation is less than 5% (high consistency), the feature weights of the fusion recognition model are fine-tuned (e.g., +0.005). If the deviation is large, the tuning model scaling factor is corrected, and the next cycle begins.

[0030] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A seismoelectric coupling advanced detection intelligent self-tuning system for TBM tunneling parameters, characterized in that, include: The seismic-electric coupling advanced detection module is installed on the cutterhead, front shield, or shield body of the TBM. It is used to conduct synchronous advanced detection of the surrounding rock ahead during the TBM tunneling process and obtain seismic response data and electrical response data in the same detection section. The multi-source data synchronous acquisition and preprocessing module is connected to the seismic-electric coupling advanced detection module. It is used to synchronize and align the seismic response data and electrical response data in time, and to perform joint correction on the seismic response data and electrical response data based on the TBM's thrust, cutterhead speed or torque tunneling condition parameters, so as to obtain seismic-electric coupling characteristic data. The geological state fusion identification and risk assessment module is connected to the multi-source data synchronous acquisition and preprocessing module. It is used to fuse and identify the structural integrity, lithological changes or water-bearing status of the surrounding rock in front of the TBM based on the seismic-electric coupling characteristic data, and output the corresponding geological state type and risk confidence evaluation results. The tunneling parameter self-tuning control module is connected to the geological state fusion identification and risk assessment module, and is used to continuously and adaptively tune the TBM's thrust, cutterhead speed, torque limit, or single-step advance tunneling parameters based on the risk confidence assessment results. The tunneling feedback and model self-updating module is connected to the tunneling parameter self-tuning control module. It is used to collect actual tunneling response data and, based on the consistency between the actual tunneling response data and the geological state fusion identification results, to self-update and correct the seismic-electric coupling feature weights, risk assessment models, or tunneling parameter tuning models, thereby forming a closed-loop TBM tunneling control system that integrates seismic-electric advanced detection, geological state identification, tunneling parameter self-tuning, and feedback correction.

2. The intelligent self-tuning system for advanced detection of TBM tunneling parameters using seismoelectric coupling as described in claim 1, characterized in that, The seismic-electric coupling advanced detection module adopts a unified detection step size and synchronous triggering mechanism, which enables the seismic wave excitation, vibration response acquisition, current injection, and electrical response acquisition to be completed within the same tunneling cycle, thereby achieving a one-to-one correspondence between seismic wave response data and electrical response data in terms of spatial location and time scale.

3. The intelligent self-tuning system for seismic-electric coupling advanced detection of TBM tunneling parameters according to claim 1, characterized in that, The seismic-electric coupling advanced detection module includes a seismic wave detection unit comprising an active excitation device and a multi-channel vibration sensor array, and an electrical detection unit comprising a flow-injection electrode and a response electrode. The seismic wave detection unit and the electrical detection unit are distributed along the circumference or axial direction of the TBM.

4. The intelligent self-tuning system for advanced detection of TBM tunneling parameters using seismoelectric coupling as described in claim 1, characterized in that, The multi-source data synchronous acquisition and preprocessing module constructs a mapping model between tunneling working parameters and seismic-electric response signals, and based on this mapping model, separates and compensates for the mechanical vibration interference component in the seismic response data and the electromagnetic coupling interference component in the electrical response data.

5. The intelligent self-tuning system for seismic-electric coupling advanced detection of TBM tunneling parameters according to claim 1, characterized in that, The geological state fusion identification and risk assessment module jointly determines the state of the surrounding rock based on the mutual verification relationship between seismic wave characteristic parameters and electrical characteristic parameters. When the seismic wave characteristics and electrical characteristics show a consistent trend, the risk confidence of the corresponding geological state is increased.

6. The intelligent self-tuning system for advanced detection of TBM tunneling parameters using seismoelectric coupling as described in claim 5, characterized in that, When the seismic wave characteristic parameters and electrical characteristic parameters show contradictory changing trends, the geological state fusion identification and risk assessment module reduces the risk confidence of the corresponding geological state or marks it as an abnormal geological state.

7. The intelligent self-tuning system for advanced detection of TBM tunneling parameters using seismoelectric coupling as described in claim 5, characterized in that, The self-tuning control module for tunneling parameters proportionally controls the adjustment range of propulsion force, cutterhead speed, or single-step advance based on the numerical change of risk confidence level, so that the tunneling parameters have a continuous functional relationship with the change of risk confidence level.

8. The intelligent self-tuning system for seismic-electric coupling advanced detection of TBM tunneling parameters according to claim 5, characterized in that, When the tunneling parameter self-tuning control module identifies a high-risk or extremely high-risk geological condition, it automatically reduces the single-step advance and limits the upper limit of torque, or triggers the corresponding safety control strategy.

9. The intelligent self-tuning system for seismic-electric coupling advanced detection of TBM tunneling parameters according to claim 5, characterized in that, The tunneling feedback and model self-updating module dynamically adjusts the weight parameters of seismic-electric coupling features in the fusion identification model based on the deviation between the actual tunneling response data and the geological condition fusion identification results.

10. The intelligent self-tuning system for seismic-electric coupling advanced detection of TBM tunneling parameters according to claim 1, characterized in that, The tunneling feedback and model self-updating module adaptively corrects the tunneling parameter self-tuning model based on changes in penetration depth, torque fluctuations, or tool wear information during the tunneling process, thereby improving the safety and efficiency of subsequent tunneling processes.

Images