Mode conversion system and method of multi-mode shield tunneling machine in composite stratum

Abstract

The invention relates to the technical field of tunneling equipment, in particular to a mode conversion system of a multi-mode shield tunneling machine in a composite stratum, which comprises a prediction layer, a sensing layer, a data bus layer, a decision-making layer and an execution layer: the prediction layer acquires front geological and hydrological information through TSP advanced geological forecast and an infrared water detector; the sensing layer integrates multiple sensors to collect tunneling parameters, and the tunneling parameters are processed through a data fusion algorithm; the data bus layer realizes data space-time alignment based on a distributed stream processing platform; the decision-making layer adopts a CNN + fuzzy logic + reinforcement learning hybrid model to generate a conversion instruction; and the execution layer drives an execution mechanism to complete EPB / SPB / TBM mode conversion, and safety is guaranteed through feedback correction. According to the method, adaptive decision making and seamless execution of mode conversion are achieved, the conversion time is remarkably shortened, tool abrasion is reduced, and the construction efficiency and safety of the composite stratum are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel boring equipment, and in particular to a mode conversion system and method for a multi-mode shield machine in a composite stratum. Background Art

[0002] As urban underground space development extends into areas with complex geological conditions, multi-mode shield machines (TBMs) have become core equipment due to their ability to adapt to complex formations such as alternating soft and hard rock, karst caves, and fault zones. Multi-mode shield machines are primarily divided into two categories: dual-mode and tri-mode. Dual-mode shield machines are further categorized as EPB / hard rock tunneling TBM dual-mode shields, EPB / slurry balance (SPB) dual-mode shields, and slurry balance (SPB) / hard rock tunneling TBM dual-mode shields. The EPB / SPB dual-mode shield combines both EPB and slurry balance pressure control modes, switching between modes by switching the slag removal system. In low-permeability formations such as clay and silt, the excavation surface is balanced by soil silo pressure. In high-permeability or high-water-pressure formations such as sand and gravel, slurry circulation is used to stabilize the excavation surface. The EPB / TBM dual-mode shield has both earth pressure balance and hard rock excavation functions. The mode conversion is achieved by replacing the cutterhead or adjusting the thrust system. The earth pressure mode is used for soft rock, weathered rock and soft soil composite formations, and the TBM mode is used for full-section hard rock formations. The SPB / TBM dual-mode shield integrates mud and water circulation and hard rock breaking functions, and is suitable for formations with alternating high water pressure hard rock and permeable soft soil. The three-mode shield machine further integrates three modes on the basis of the dual mode to achieve fully adaptable excavation in extremely complex formations, and is compatible with mud circulation, spiral slag removal and roller cutter rock breaking systems in a limited space. However, the existing technology mainly has the following defects:

[0003] 1. The mode conversion of traditional multi-mode shield machines relies on manual experience judgment of preset geological parameters or offline simulation, which cannot adapt to the dynamic changes of complex strata and has the risk of response lag and misjudgment. Statistics show that the manual misjudgment rate is as high as 18%; 2. Although the existing online multi-mode shield machines have shortened the conversion time, the slag discharge system layout is fixed and lacks geological adaptability, and it is very easy to cause abnormal wear of the tool; 3. The mode conversion decision is not closed-loop: grouting reinforcement is used to ensure conversion safety, but it is not linked to real-time formation parameters, and there is a risk of lag; 4. The mechanical system has poor coordination during the mode conversion process, and there are safety hazards such as hydraulic shock and sealing failure.

[0004] Therefore, there is an urgent need to develop a full-link adaptive system that integrates multi-source sensing, dynamic decision-making and mechanical linkage to solve the real-time and safety issues of mode conversion in complex formations. Summary of the Invention

[0005] In order to solve the above technical problems, the present invention specifically adopts the following technical solutions.

[0006] Design a mode conversion system for a multi-mode shield machine in composite strata, including

[0007] Prediction layer: includes an advance detection module, which is used to obtain geological structure characteristics and hydrological information within a preset range in front of the shield machine;

[0008] Perception layer: includes a real-time monitoring module, which is used to monitor and collect operating parameters and working condition data of the shield machine during excavation, and process the real-time monitoring data through data fusion algorithms;

[0009] Data bus layer: Based on a distributed stream processing platform, it aligns and fuses the advance detection data of the prediction layer with the real-time monitoring data of the perception layer in the time and space dimensions;

[0010] Decision-making layer: Based on the time-space aligned advance detection data and real-time monitoring data, a hybrid model integrating convolutional neural network (CNN) geological feature extraction, fuzzy logic control and reinforcement learning is used to generate mode conversion decision instructions;

[0011] Execution layer: responds to decision-making instructions, drives the shield machine actuator to complete the conversion between multiple modes, and dynamically adjusts the execution parameters through closed-loop feedback.

[0012] Preferably, the advanced detection module includes an advanced geological prediction device and an infrared water detector, the advanced geological prediction device can detect at least 100m ahead, and the infrared water detector can identify the fault water content information within a range of at least 30m ahead.

[0013] Preferably, the real-time monitoring module includes a cutter head torque sensor, a propulsion speed sensor, a soil bin pressure sensor and a water seepage monitor;

[0014] The data fusion algorithm includes a wavelet transform denoising algorithm and a Kalman filter algorithm.

[0015] The wavelet transform denoising algorithm adaptively decomposes the sensor signal through wavelet layered denoising, and the number of decomposition layers is dynamically selected according to the signal bandwidth;

[0016] The Kalman filter algorithm fuses multi-sensor data and introduces sensor reliability weight factors for state estimation;

[0017] An online compensation model for drift error of soil bin pressure sensor is established.

[0018] Preferably, the data bus layer uses a distributed stream processing platform to implement data integration. The distributed stream processing platform is the Apache Kafka platform, and data alignment is achieved through the following methods:

[0019] Data standardization: Protobuf protocol is used to process real-time monitoring data, and Avro+GeoJSON protocol is used to process advanced detection data;

[0020] Time alignment: The NTP server is used to unify the time of all devices, with a time error of less than 1ms;

[0021] Spatial alignment: Establish the transformation relationship between the shield machine's local coordinate system and the project's global coordinate system to achieve spatial position matching between monitoring data and geological data.

[0022] Preferably, the algorithm of the hybrid model includes:

[0023] Geological feature extraction model: A dual-branch processing structure is used to process advanced geological exploration image data and water-bearing identification parameters respectively, and output stratum classification results, permeability coefficients, and risk area information;

[0024] Fuzzy logic control: A dynamic rule base is built based on denoised real-time monitoring data, including cutterhead torque, propulsion speed, soil bin pressure, and water seepage.

[0025] Reinforcement learning: Taking surface subsidence increment, energy consumption, and equipment health as optimization objectives, the control strategy is adjusted through the reward function, and mode conversion instructions and execution parameters are output.

[0026] Preferably, the execution layer includes an instruction parsing module, an action execution module, a security monitoring module, and a feedback correction module;

[0027] in,

[0028] Instruction parsing module, used to decompose decision instructions into multi-system coordinated action sequences;

[0029] The action execution module is used to drive the coordinated actions of the hydraulic system, grouting system, and soil discharge system, and realize the conversion between earth pressure balance mode, slurry balance mode, and hard rock excavation mode;

[0030] Security monitoring module, used to verify execution status in real time and trigger abnormal rollback or emergency shutdown;

[0031] The feedback correction module dynamically adjusts execution parameters based on sensor closed-loop feedback.

[0032] Preferably, the advanced geological prediction equipment uses seismic wave method to detect the structural status and integrity of the surrounding rock, and the infrared water detector identifies hidden water sources by detecting the distortion of the infrared radiation field of the geological body. The two match structural abnormality areas and water-bearing abnormality areas through a complementary data fusion mechanism.

[0033] Preferably, in the dual-branch processing structure of the geological feature extraction model, one branch processes seismic wave image data and the other processes infrared parameters; the reinforcement learning training process is combined with expert experience data, and the action exploration range is limited within the engineering safety threshold.

[0034] Preferably, the mode conversion process includes the coordinated actions of screw conveyor locking, mud pump pressure regulation and shield tail seal dynamic control; the abnormal rollback strategy of the safety monitoring module is dynamically triggered according to the seepage volume and soil bin pressure deviation.

[0035] A method for mode conversion of a multi-mode shield machine in a composite stratum comprises the following steps:

[0036] Step S1: obtaining geological structure characteristics and hydrological information within a preset range in front of the shield machine through advanced detection equipment;

[0037] Step S2: collecting operating parameters and working condition data of the shield machine during the tunneling process through multiple real-time monitoring sensors, and processing the operating parameters and working condition data using a data fusion algorithm;

[0038] Step S3: Based on the distributed stream processing platform, the geological and hydrological data obtained in step S1 and the real-time monitoring data processed in step S2 are aligned and integrated in the time dimension and the spatial dimension;

[0039] Step S4: Based on the aligned and fused data in step S3, a hybrid algorithm of geological feature extraction model, fuzzy logic control and reinforcement learning is used to generate mode conversion decision instructions and corresponding execution parameters;

[0040] Step S5: According to the mode conversion decision instruction, the shield machine's actuator is driven to complete the conversion between the earth pressure balance mode, the slurry balance mode and the hard rock tunneling mode, and the execution parameters are dynamically adjusted through the closed-loop feedback of the sensor. At the same time, the conversion process is safely monitored and an emergency response is triggered in abnormal situations.

[0041] The beneficial effects of the present invention are:

[0042] 1. This invention uses the TSP advanced geological prediction equipment and infrared water detector in the prediction layer to achieve advanced detection of geological structures 100 meters ahead and the water content of faults 30 meters below. Combined with real-time monitoring by sensors such as cutterhead torque, propulsion speed, and soil bin pressure in the perception layer, and precise spatial and temporal alignment in the data bus layer, it provides comprehensive data support for decision-making. A hybrid model integrating CNN geological feature extraction and reinforcement learning can dynamically identify stratum changes and generate optimal conversion strategies, shortening conversion time from the traditional 35 minutes to less than 18 minutes.

[0043] 2. The present invention adopts a data fusion algorithm of Kalman filtering and wavelet transform to effectively improve the accuracy of monitoring data and reduce misjudgments caused by data noise; the decision-making mechanism combining fuzzy logic and reinforcement learning can dynamically adjust parameters such as cutterhead torque and soil bin pressure according to the characteristics of the formation, reducing the tool wear rate by more than 42%; the safety monitoring and feedback correction module of the execution layer controls the surface settlement to 3.2mm (lower than the standard value of 8mm) through abnormal rollback and emergency shutdown mechanisms, significantly improving construction safety.

[0044] 3. The present invention constructs a full-process adaptive closed-loop control system of "advanced prediction-real-time monitoring-intelligent decision-making-precise execution", breaking through the traditional reliance on manual experience. Through multi-source data fusion and intelligent algorithm, it autonomously optimizes the mode conversion strategy to adapt to the dynamic changes of complex formations; at the same time, the coordinated linkage of modular actuators (such as screw conveyor locking, mud pump pressure regulation, etc.) realizes seamless conversion between EPB / SPB / TBM modes, thereby improving construction efficiency.

[0045] 4. The mode conversion system of the present invention can accurately identify different strata types such as clay, sand, hard rock, and water-rich risk areas, dynamically match the optimal excavation mode, and perform well in extremely complex strata such as water-rich faults and alternating soft and hard rocks. It not only reduces energy consumption (for example, the energy consumption of the mud circulation system is reduced to 1.8kW·h / m 3 , 33% less than traditional methods), and also reduces rework and equipment maintenance costs, with significant engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the overall system architecture of the present invention;

[0047] Figure 2 It is a schematic diagram of the data fusion processing process;

[0048] Figure 3 It is a data processing flow chart of the data bus layer;

[0049] Figure 4 It is a collaborative decision-making flowchart;

[0050] Figure 5 It is the execution layer processing diagram; DETAILED DESCRIPTION

[0051] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0052] Example 1

[0053] A mode conversion system for a multi-mode shield machine in a composite stratum, such as Figures 1 to 5 Shown, including

[0054] Prediction layer: includes an advance detection module, which is used to obtain geological structure characteristics and hydrological information within a preset range in front of the shield machine; the advance detection module includes an advance geological prediction device and an infrared water detector. The advance geological prediction device can detect at least 100m in front, and the infrared water detector can identify fault water content information within a range of at least 30m in front. The advance geological prediction device uses the seismic wave method to detect the state and integrity of the surrounding rock structure. The model is TSP-SK. The infrared water detector uses the HW-304 tunnel infrared water detector, which identifies hidden water sources by detecting the distortion of the infrared radiation field of the geological body. The resolution is: H mode 0.05mW / cm 2 , M mode 0.07mW / cm 2 ; Working temperature: 0-40℃, humidity ≤80%. The two match structural abnormal areas and water abnormal areas through a complementary data fusion mechanism.

[0055] Perception layer: includes a real-time monitoring module, which is used to monitor and collect the operating parameters and working condition data of the shield machine during the excavation process in real time, and process the real-time monitoring data through a data fusion algorithm; the real-time monitoring module includes a cutter head torque sensor, a propulsion speed sensor, a soil bin pressure sensor and a water seepage monitor; the data fusion algorithm includes a wavelet transform denoising algorithm and a Kalman filter algorithm. Wavelet transform is a time-frequency analysis tool that performs multi-resolution decomposition of signals through basis functions of different scales, and can accurately locate the frequency band distribution of noise and effective signals.

[0056] Kalman filtering is mainly used to denoise and estimate the state of real-time sensor data, and correct sensor drift to ensure the immediate accuracy of the control algorithm input data. For example, when a sudden increase in torque is detected, it immediately triggers emergency deceleration, which can significantly improve data quality and achieve efficient fusion and state estimation of multi-source sensors.

[0057] The wavelet transform denoising algorithm adaptively decomposes the sensor signal through wavelet layered denoising, with the number of decomposition layers dynamically selected based on the signal bandwidth. The Kalman filter algorithm fuses multi-sensor data and introduces sensor reliability weighting factors for state estimation. An online drift error compensation model for the soil silo pressure sensor is established. Specifically, it includes:

[0058] Time synchronization: All sensors are equipped with GPS synchronization clock modules, with timestamp accuracy ≤ 1ms;

[0059] Spatial mapping: Establish the conversion relationship between the shield machine coordinate system (local) and the project global coordinate system:

[0060]

[0061] Where S is the excavation mileage and θ is the shield machine attitude angle.

[0062] Wavelet layered denoising, adaptive selection of decomposition layers, dynamic determination of decomposition layer number J according to signal bandwidth:

[0063]

[0064] Among them, fs is the sampling frequency and fc is the cutoff frequency of the effective component of the signal.

[0065] Improved threshold function: Use segmented adaptive threshold method to process high frequency coefficients.

[0066]

[0067] Among them, σ j is the estimated value of the noise standard deviation of the jth layer, and N is the signal length.

[0068] Multi-sensor Kalman fusion, state vector design,

[0069] X k =[T k ,P k ,V k ,Q k ] T

[0070] Among them, T represents torque, P represents soil bin pressure, V represents propulsion speed, and Q represents the optimal estimated value of water seepage.

[0071] Improve the observation model and introduce the sensor reliability weight factor α i :

[0072]

[0073] Among them, z k represents the observed and measured values ​​at time point k; H k represents the observation matrix or measurement matrix at time point k; x k is the system state vector at time k; Indicates T k ,P k ,V k ,Q k 4 noises; v i,k represents the noise of the i-th sensor at time point k; α i Dynamically adjusted based on the sensor's historical failure rate.

[0074] Drift error online compensation, drift model: establish the drift compensation equation for the soil bin pressure sensor:

[0075] P corrected =P raw -β·t

[0076] Among them, P corrected Indicates the pressure value after compensation; P raw represents the original pressure value; β represents the drift coefficient, which is estimated online by the sliding window least squares method; t represents the time interval between the current time and the calibration time.

[0077] Data bus layer: Based on a distributed stream processing platform, the advanced detection data of the prediction layer and the real-time monitoring data of the perception layer are aligned and integrated in the time and space dimensions. The data bus layer uses a distributed stream processing platform for data integration. The distributed stream processing platform is the Apache Kafka platform, which supports high-throughput, low-latency real-time data transmission and is commonly used to build data pipelines and message buses. Data alignment is achieved through the following methods:

[0078] Data standardization: Protobuf protocol is used to process real-time monitoring data, and Avro+GeoJSON protocol is used to process advanced detection data;

[0079] Time alignment: The NTP server is used to unify the time of all devices, with a time error of less than 1ms. High-precision timestamps are embedded in the data header for time alignment.

[0080] Spatial alignment: Establish the transformation relationship between the shield machine's local coordinate system and the project's global coordinate system to achieve spatial position matching between monitoring data and geological data.

[0081] Specifically, the spatial coordinate mapping relationship during spatial alignment is as follows: the real-time sensor position (shield machine relative coordinate system) is converted into the engineering absolute coordinate system for real-time positioning through the shield machine guidance system; the advanced detection of geological features (geographic coordinate system WGS84) is converted into the engineering coordinate system through the Proj4 library.

[0082] Based on the Apache Kafka platform, the spatiotemporal consistency fusion of real-time monitoring data and advanced detection data is achieved through spatiotemporal composite key partitioning, stream processing window alignment and dynamic coordinate mapping.

[0083] The spatiotemporal composite key includes the ring number, timestamp, and geological feature type, and the partition allocation strategy is dynamically calculated based on the ring number hash value.

[0084] The stream processing window is configured with a sliding time window and a watermark mechanism to tolerate data out-of-order delays.

[0085] Use Kafka Streams state storage and store geological features in a geospatial database to enable real-time queries in stream processing.

[0086] The Kafka cluster optimization can be partitioned according to the shield machine ID or ring number to ensure that data in the same spatial location is in the same partition.

[0087] Decision-making layer: Based on the spatiotemporally aligned advance detection data and real-time monitoring data, a hybrid model integrating convolutional neural network (CNN) geological feature extraction, fuzzy logic control, and reinforcement learning is used to generate mode conversion decision instructions. The algorithm of the hybrid model includes:

[0088] Geological feature extraction model: This model extracts multimodal geological features by parsing aligned seismic waveforms and infrared thermal imaging data to extract features such as stratum classification and permeability.

[0089] Input advanced geological prediction images (TSP seismic wave method waveforms) and infrared water exploration parameters (temperature gradient matrix (ΔT / Δx) output by infrared thermal imagers) data.

[0090] A dual-branch processing structure is used to process advanced geological exploration image data and water-bearing identification parameters respectively, outputting stratum classification results, permeability coefficients, and risk area information. The output stratum classification is clay, sand, and hard rock.

[0091] Output permeability coefficient regression k value (×10 -5 m / s), segmenting the risk area and identifying the boundaries of water-rich fractures and cavities; and identifying the state of the tunnel face ahead by analyzing relevant data. The dual-branch processing structure of the geological feature extraction model processes seismic wave image data on one branch and infrared parameters on the other.

[0092] Fuzzy logic control: A dynamic rule base is constructed based on de-noised real-time monitoring data, including cutterhead torque T (range 0-2500 kN·m), propulsion speed V (range 0-60 mm / min), soil bin pressure P (range 0-0.6 MPa), and water seepage Q (range 0-50 L / min).

[0093] Among them, fuzzy logic control converts the current precise perception layer related data into fuzzy language values, then defines the mapping relationship between numerical values ​​and fuzzy sets, calculates the output fuzzy sets based on the input fuzzy values ​​and rule base, and then defuzzifies and converts the fuzzy output into precise control quantities.

[0094] Among them, the cutterhead torque is directly affected by the hardness, density and rock-soil friction characteristics of the formation, and can be used as one of the important bases for judging the timing of mode switching of a multi-mode shield machine. When the torque is continuously monitored to exceed the current mode threshold and lasts for a certain period of time, the system can issue an early warning or automatically trigger the mode switching logic.

[0095] Among them, the advancing speed is a direct feedback of the formation hardness. For example, the advancing speed is usually higher in soft soil formations, and the advancing speed drops significantly in hard rock or composite formations.

[0096] Among them, the soil bin pressure needs to be dynamically matched with the stratum water and soil pressure to prevent the excavation surface from becoming unstable. It can be used as one of the core parameters for judging the mode conversion of multi-mode shield machines.

[0097] Among them, the water seepage parameter is mainly reflected in the real-time feedback of the formation permeability and groundwater status and the prediction of construction risks. Low water seepage indicates that the formation permeability is low, mostly clay, silty clay, etc. High water seepage indicates that the formation permeability is high, mostly sand layers, pebble layers, etc.

[0098] Fuzzy design: Use asymmetric Gaussian function to optimize the membership function to adapt to the dynamic range of parameters.

[0099]

[0100] Among them, x is the input variable, which represents the original parameters that need to be fuzzified (cutter head torque, propulsion speed, soil bin pressure, and water seepage); σ L and σ R is the left and right standard deviation (σ L =σ R ), which controls the steepness of the curves on both sides of the center point; c is the center point of the membership function, and t is the segmentation threshold, which divides the application interval of the left and right standard deviations.

[0101] Dynamic rule base: Set 25 initial rules and optimize weights online through reinforcement learning.

[0102] Reinforcement learning: Optimizing ground subsidence increments, energy consumption, and equipment health, the system uses a reward function to adjust control strategies and output mode transition instructions and execution parameters. The reinforcement learning training process incorporates expert experience data, and the range of action exploration is limited to engineering safety thresholds.

[0103] The reinforcement learning strategy optimization is specifically as follows:

[0104] State space: s t =[CNN feature vector (32 dimensions), fuzzification parameter (6 dimensions), historical action sequence (5 steps)]

[0105] Action space: Discrete action: {EPB mode, SPB mode, TBM mode}

[0106] Continuous action: mud pressure setting value (0.2~0.6MPa), foam injection rate (0~50L / min).

[0107] Step S47, setting the reward function:

[0108]

[0109] Among them, R t represents the reward value at time t, V t Indicates the propulsion speed, V max Indicates the maximum propulsion speed, ΔS t is the surface settlement increment, S th is the sedimentation threshold, E t is energy consumption, E max is the maximum allowable energy consumption, H t The equipment status is scored (0-100%), and the shield maintenance personnel will judge and score it every day, with 100% indicating the best status.

[0110] Training algorithm: Proximal Policy Optimization (PPO) combined with expert experience playback, with expert data accounting for 30% (historical operation records), and the action exploration range is limited to the engineering safety threshold of ±10%.

[0111] Multimodal decision fusion

[0112] Emergency layer: Fuzzy logic directly responds to sudden increases in water seepage (>30L / min) or torque exceeding the limit (>2000kN·m), triggering mode conversion;

[0113] Optimization layer: Reinforcement learning generates long-term strategies (e.g., planning tool changes 50 rings in advance);

[0114] Human-machine collaboration: Under abnormal working conditions, instructions with a confidence level of <80% are pushed to manual confirmation.

[0115] Execution layer: responds to decision-making instructions, drives the shield machine actuator to complete the transition between multiple modes, and dynamically adjusts the execution parameters through closed-loop feedback. The execution layer includes the instruction parsing module, action execution module, safety monitoring module, and feedback correction module;

[0116] Among them, the instruction parsing module is used to decompose the decision instructions (mode conversion, parameter adjustment) into a multi-system collaborative action sequence;

[0117] The action execution module is used to drive the coordinated actions of the hydraulic system, grouting system, and soil discharge system, and realize the conversion between earth pressure balance mode, mud-water balance mode, and hard rock excavation mode; the mode conversion process includes the coordinated actions of screw conveyor locking, mud pump pressure regulation, and shield tail seal dynamic control;

[0118] The safety monitoring module verifies execution status in real time and triggers abnormal rollback or emergency shutdown. The feedback correction module dynamically adjusts execution parameters based on closed-loop sensor feedback. The safety monitoring module's abnormal rollback strategy is dynamically triggered based on water seepage and soil silo pressure deviation.

[0119] Example 2

[0120] A method for mode conversion of a multi-mode shield machine in a composite stratum comprises the following steps:

[0121] Step S2: collecting operating parameters and working condition data of the shield machine during the tunneling process through multiple real-time monitoring sensors, and processing the operating parameters and working condition data using a data fusion algorithm;

[0122] Step S3: Based on the distributed stream processing platform, the geological and hydrological data obtained in step S1 and the real-time monitoring data processed in step S2 are aligned and integrated in the time dimension and the spatial dimension;

[0123] Step S4: Based on the aligned and fused data in step S3, a hybrid algorithm of geological feature extraction model, fuzzy logic control and reinforcement learning is used to generate mode conversion decision instructions and corresponding execution parameters;

[0124] Step S5: According to the mode conversion decision instruction, the shield machine's actuator is driven to complete the conversion between the earth pressure balance mode, the slurry balance mode and the hard rock tunneling mode, and the execution parameters are dynamically adjusted through the closed-loop feedback of the sensor. At the same time, the conversion process is safely monitored and an emergency response is triggered in abnormal situations.

[0125] The present invention switches between earth pressure balance (EPB), slurry balance (SPB), and hard rock tunneling (TBM) modes:

[0126] 1. Conversion from Earth Pressure Balance (EPB) to Slurry Water Balance (SPB) Mode

[0127] Application scenarios:

[0128] When the shield machine excavated to mileage SK12+356, the advance detection module detected a high permeability sand layer 30m ahead (permeability coefficient k = 2.3×10 -5 m / s), the TSP waveform showed that the formation integrity coefficient dropped sharply from 0.85 to 0.52, and the infrared water detector detected that the water content in the fault zone was 18%.

[0129] Implementation steps:

[0130] Step S1, data trigger (timestamp 2023-08-15T09:23:15.356Z)

[0131] Real-time data from the sensor layer: The cutterhead torque suddenly increased to 1850 kN·m (EPB mode threshold: 1800 kN·m), the soil bin pressure fluctuated by ±0.12 MPa (exceeding the limit by ±0.05 MPa), and the water seepage rate increased from 8 L / min to 32 L / min.

[0132] Step S2, data fusion: after wavelet denoising, J=5-layer decomposition, Kalman filter compensation soil bin pressure drift β=0.003MPa / h;

[0133] Step S3, decision generation (response time < 800ms);

[0134] CNN output: Sand layer probability 92%, permeability coefficient k = 2.1×10 -4 m / s (error <8%);

[0135] Reinforcement learning strategy: reward function Rt = 0.72 (historical optimal Rmax = 0.81), generating action sequence:

[0136]

[0137] Step S4, execution control: screw conveyor locked (time consumption 2min15s, pressure gradient 0.05MPa / min),

[0138] Mud pump group starts (flow Q = 420m 3 / h, pipe diameter Φ250mm), shield tail seal dynamic adjustment (injection of sealing grease 1.2kg / m, pressure 0.25MPa);

[0139] Step S5, safety monitoring: soil bin pressure closed loop control: PID parameters Kp = 2.5, Ki = 0.8, Kd = 0.3,

[0140] Abnormal rollback triggering conditions: switch back to EPB mode when the water seepage volume is greater than 45L / min for 30s.

[0141] Technical effect: The mode conversion time is shortened from the traditional 35 minutes to 18 minutes, the tool wear rate is reduced by 42%, and the surface settlement is controlled at 3.2mm (8mm lower than the standard value). The reduction ratio of the tool wear rate is obtained from the following "Comparison of the wear rate of the roller cutter before and after the mode conversion": Specifically, the arithmetic average of the reduction ratio of the volume wear rate index of 62 tools is calculated, and it is finally concluded that after the mode conversion, the tool wear rate is reduced by about 42%. Specifically, the volume wear rate of the roller cutter refers to the radial wear amount generated by the roller cutter excavating a unit volume of rock. The calculation method is as follows: i =Q i / π(R i 2 -R i-1 2 )L, where v i is the rock breaking volume wear rate of the i-th cutter on the cutterhead, in mm / m 3 ;Q i is the cumulative wear of the i-th hob, in mm; R iis the installation radius of the i-th hob, in m; R i-1 is the installation radius of the previous cutter, in meters; L is the shield's tunneling distance, in meters. The greater wear on cutters 59#, 61#, and 62# is primarily due to more severe wear on the edge cutters. The smaller wear on cutter 61# is likely due to manual measurement errors.

[0142]

[0143]

[0144] 2. Transition from Hard Rock Tunneling (TBM) to Slurry Balance (SPB) Mode

[0145] Application scenarios:

[0146] When the tunnel reached mileage SK14+215 and crossed a water-rich fault zone, the TSP showed that the RQD value of the rock mass 50m ahead dropped sharply from 85% to 32%. The infrared water detector detected a water temperature gradient of ΔT / Δx = 0.8°C / m (threshold 0.5°C / m).

[0147] Implementation steps:

[0148] Step S1, data fusion: advanced detection data: fault width 4.8m, permeability coefficient k = 5.6×10 -5 m / s,

[0149] Real-time sensor: The cutterhead vibration frequency increased from 58Hz to 82Hz, and the propulsion speed decreased from 18mm / min to 6mm / min;

[0150] Step S2, decision optimization: CNN feature vector: [0.87, 0.12, 0.01] (hard rock / fault breccia / clay probability distribution), fuzzy rule activation: Rule #17 (high vibration + low propulsion speed → mud wall requirement 0.89), reinforcement learning output: mud viscosity setting 22s (Madhyama funnel), injection pressure 0.42 MPa;

[0151] Step S3, execution process:

[0152] Cutter mode switching: the retraction angle of the hob is adjusted from 75° to 30°, and the tool load is balanced;

[0153] Slag discharge system conversion: belt conveyor stop → mud circulation system start (takes 9 minutes and 23 seconds);

[0154] Synchronous grouting parameters: water-cement ratio 0.45:1, grouting pressure 0.28MPa (35% higher than normal);

[0155] Step S4, safety protection: dynamic limit of cutter head torque: the maximum value is limited to 85% of the rated torque (2200 kN·m);

[0156] Emergency grouting system pre-start: reserve slurry 30m 3 Always on call.

[0157] Technical effect: The energy consumption of the mud circulation system is 2.7kW·h / m compared with the traditional method. 3 Reduced to 1.8kW·h / m 3 .

[0158] 3. Transition from Hard Rock Tunneling (TBM) to Earth Pressure Balance (EPB) Mode

[0159] Application scenarios:

[0160] When excavating to mileage SK15+780, advance detection showed that a clay interlayer (2.3m thick, 24% water content) appeared 18m ahead, and the cutterhead torque in TBM mode remained below 800kN·m.

[0161] Implementation steps:

[0162] Step S1, data preprocessing: wavelet denoising: using the db8 wavelet basis, decomposition level J = 4, the signal-to-noise ratio is improved to 38 dB; Kalman fusion: state vector X_k = [785 kN·m, 0.11 MPa, 52 mm / min, 5 L / min]^T;

[0163] Step S2, hybrid decision: CNN multi-task output: clay probability 89%, permeability coefficient k = 3.2×10 -6 m / s;

[0164] Fuzzy control: membership μ_T(785) = 0.67 (TBM mode mismatch > 60%); reinforcement learning reward: R_t = 0.65 (mainly from the propulsion speed gain term);

[0165] Step S3, actuator linkage: cutterhead conversion: full cross-section of the hob is retracted (taking 6min10s), and the scraper extension is increased to 120mm; soil bin pressure is established: the screw conveyor speed is accelerated from 0rpm to 8rpm (gradient 0.5rpm / s 2 ); Foam system parameters: injection rate 12L / min, expansion ratio 25:1;

[0166] Step S4, dynamic feedback: earth pressure balance control: pressure setting value P_set = 0.18 MPa (dynamically calculated according to the burial depth); correction mechanism: detect the cutter head torque deviation every 30 seconds, and trigger parameter re-optimization if it exceeds ±5%.

[0167] Technical Effect: The time required to establish soil silo pressure during the mode conversion process is shortened from 8 minutes in the traditional method to 3 minutes. Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A mode conversion system for a multi-mode shield machine in a composite stratum, characterized in that: Including prediction layer: including advance detection module, used to obtain geological structure characteristics and hydrological information within the preset range in front of the shield machine; Perception layer: includes a real-time monitoring module, which is used to monitor and collect operating parameters and working condition data of the shield machine during excavation, and process the real-time monitoring data through data fusion algorithms; Data bus layer: Based on a distributed stream processing platform, it aligns and fuses the advance detection data of the prediction layer with the real-time monitoring data of the perception layer in the time and space dimensions. Decision-making layer: Based on the time-space aligned advance detection data and real-time monitoring data, a hybrid model integrating convolutional neural network (CNN) geological feature extraction, fuzzy logic control and reinforcement learning is used to generate mode conversion decision instructions; Execution layer: responds to decision-making instructions, drives the shield machine actuator to complete the conversion between multiple modes, and dynamically adjusts the execution parameters through closed-loop feedback.

2. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 1, characterized in that: The advanced detection module includes an advanced geological prediction device and an infrared water detector. The advanced geological prediction device can detect at least 100 meters ahead, and the infrared water detector can identify the fault water content information within a range of at least 30 meters ahead.

3. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 1, characterized in that: The real-time monitoring module includes a cutter head torque sensor, a propulsion speed sensor, a soil bin pressure sensor and a water seepage monitor; The data fusion algorithm includes a wavelet transform denoising algorithm and a Kalman filter algorithm. The wavelet transform denoising algorithm adaptively decomposes the sensor signal through wavelet layered denoising, and the number of decomposition layers is dynamically selected according to the signal bandwidth; The Kalman filter algorithm fuses multi-sensor data and introduces a sensor reliability weight factor for state estimation; an online compensation model for the drift error of the soil bin pressure sensor is established.

4. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 1, characterized in that: The data bus layer uses a distributed stream processing platform to achieve data integration. The distributed stream processing platform is the Apache Kafka platform, and data alignment is achieved through the following methods: Data standardization: Protobuf protocol is used to process real-time monitoring data, and Avro+GeoJSON protocol is used to process advanced detection data; Time alignment: The NTP server is used to unify the time of all devices, with a time error of less than 1ms; Spatial alignment: Establish the transformation relationship between the shield machine's local coordinate system and the project's global coordinate system to achieve spatial position matching between monitoring data and geological data.

5. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 1, characterized in that: The algorithm of the hybrid model includes: Geological feature extraction model: A dual-branch processing structure is used to process advanced geological exploration image data and water-bearing identification parameters respectively, and output stratum classification results, permeability coefficients, and risk area information; Fuzzy logic control: A dynamic rule base is built based on denoised real-time monitoring data, including cutterhead torque, propulsion speed, soil bin pressure, and water seepage. Reinforcement learning: Taking surface subsidence increment, energy consumption, and equipment health as optimization objectives, the control strategy is adjusted through the reward function, and mode conversion instructions and execution parameters are output.

6. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 1, characterized in that: The execution layer includes an instruction parsing module, an action execution module, a security monitoring module, and a feedback correction module; in, Instruction parsing module, used to decompose decision instructions into multi-system coordinated action sequences; The action execution module is used to drive the coordinated actions of the hydraulic system, grouting system, and soil discharge system, and realize the conversion between earth pressure balance mode, slurry balance mode, and hard rock excavation mode; Security monitoring module, used to verify execution status in real time and trigger abnormal rollback or emergency shutdown; The feedback correction module dynamically adjusts execution parameters based on sensor closed-loop feedback.

7. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 2, characterized in that: The advanced geological prediction equipment uses seismic wave method to detect the structural status and integrity of the surrounding rock, and the infrared water detector identifies hidden water sources by detecting the distortion of the infrared radiation field of the geological body. The two match structural abnormality areas and water-bearing abnormality areas through a complementary data fusion mechanism.

8. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 5, characterized in that: In the dual-branch processing structure of the geological feature extraction model, one branch processes seismic wave image data and the other processes infrared parameters; The reinforcement learning training process is combined with expert experience data, and the range of action exploration is limited to the engineering safety threshold.

9. The mode conversion system for a multi-mode shield machine in a composite stratum according to claim 6, characterized in that: The conversion process between the earth pressure balance mode, slurry balance mode and hard rock tunneling mode includes the coordinated actions of screw conveyor locking, mud pump pressure regulation and shield tail seal dynamic control; the abnormal rollback strategy of the safety monitoring module is dynamically triggered according to the seepage volume and soil bin pressure deviation.

10. A mode conversion method for a multi-mode shield machine in a composite stratum, which is implemented based on the mode conversion system for a multi-mode shield machine in a composite stratum according to any one of claims 1 to 9, and is characterized by: The following steps are involved: Step S1: obtaining geological structure characteristics and hydrological information within a preset range in front of the shield machine through advanced detection equipment; Step S2: collecting operating parameters and working condition data of the shield machine during the tunneling process through multiple real-time monitoring sensors, and processing the operating parameters and working condition data using a data fusion algorithm; Step S3: Based on the distributed stream processing platform, the geological and hydrological data obtained in step S1 and the real-time monitoring data processed in step S2 are aligned and integrated in the time dimension and the spatial dimension; Step S4: Based on the aligned and fused data in step S3, a hybrid algorithm of geological feature extraction model, fuzzy logic control and reinforcement learning is used to generate mode conversion decision instructions and corresponding execution parameters; Step S5: According to the mode conversion decision instruction, the shield machine's actuator is driven to complete the conversion between the earth pressure balance mode, the slurry balance mode and the hard rock tunneling mode, and the execution parameters are dynamically adjusted through the closed-loop feedback of the sensor. At the same time, the conversion process is safely monitored and an emergency response is triggered in abnormal situations.

Images