Dynamic tunnel face pressure balancing system and method based on multi-agent collaborative decision

Abstract

The invention belongs to the technical field of shield tunnels, and discloses a tunnel face pressure dynamic balance system and method based on multi-agent collaborative decision-making. Real-time self-adaptive regulation and control are achieved; the system can automatically adjust the tunneling speed and the tunnel face pressure in real time according to real-time monitoring data and a tunnel face stability judgment result; real-time self-adaption control is achieved, complex and changeable stratum conditions are effectively dealt with, and construction safety and stability are improved. According to the effective treatment of special working conditions, a reasonable control strategy can be formulated, the pressure of a tunnel face is stably transited, settlement is strictly controlled, and the construction safety and the surrounding environment safety are guaranteed according to special application scenes such as complex composite stratum crossing and extremely close distance underneath crossing of existing buildings and structures and abnormal tunneling states such as shield starting, stopping and speed changing.

Description

Technical Field

[0001] This invention belongs to the field of shield tunnel technology, and particularly relates to a dynamic pressure balancing system and method for the tunnel face based on multi-agent collaborative decision-making. Background Technology

[0002] Three representative prior art patents or published documents that are relatively close to the subject matter of this invention (pressure control system at the tunnel face) are cited, and their respective remaining technical problems are pointed out to further emphasize the innovative points of this invention:

[0003] 1. US4167289A ("Method and system for controlling earth pressure intunnel boring or shield machine")

[0004] This patent proposes a method to maintain the pressure within a predetermined range by measuring the earth pressure in the tunnel boring machine chamber or at the working face, comparing it with a reference value, and then controlling the advance speed or the amount of soil transported.

[0005] The existing technical problems include: the system mainly uses single pressure measurement and propulsion / slag removal as control means, lacks simultaneous dynamic monitoring and judgment of multiple parameters (such as geological conditions, support strength, settlement data, etc.); and it does not build a stability discrimination module, fuzzy logic or machine learning model to comprehensively analyze the stress and deformation state of the working face.

[0006] Furthermore, it lacks specialized control strategies or module designs for complex working conditions (such as soft upper and hard lower strata, and extremely close crossing of existing tunnels), making it difficult to meet the stable control requirements under high-risk complex strata.

[0007] 2. CN108414128B (“Device and method for monitoring earth pressure onexcavation surface of model shield machine”)

[0008] This open-source document provides a device and method for monitoring earth pressure at the excavation face of a model tunnel boring machine, focusing on technical improvements in monitoring aspects such as sensor placement and wireless data acquisition.

[0009] While it solves the problem of real-time monitoring, its technical limitations lie in its limitation to model test scenarios and its focus on the pressure monitoring device itself, without further transforming the monitoring results into automated control strategies. In other words, it does not establish a complete "discrimination → control" closed loop, nor does it incorporate multi-source data (settlement, support status, geological scores, etc.) into the discrimination or control logic.

[0010] Meanwhile, this technology does not consider dedicated control modules or switching strategies for complex working conditions, thus limiting its application in real construction environments, especially in complex geological formations or special tunnel crossing scenarios.

[0011] 3. CN102410027A (“Novel dual-mode TBM and control method thereof”)

[0012] This patent proposes a shield tunneling machine and control method with switchable modes (soft soil EPB mode / hard rock TBM mode) to broaden the equipment's adaptability to different geological formations.

[0013] The technical problem lies in its focus on the mode switching capability of the tunneling equipment itself, rather than on intelligent control of multiple parameters such as face pressure, tunneling speed, and surface settlement. Furthermore, while its control method improves equipment adaptability, it lacks a stability discrimination module based on fuzzy logic or geological discriminant functions, and it does not utilize monitoring data feedback for real-time adjustment of tunneling speed / pressure through automated closed-loop control.

[0014] In certain complex scenarios (such as passing under an operational tunnel at extremely close range, or in areas with a soft upper stratum and a hard lower stratum), the lack of a targeted "special working condition processing module" design may result in insufficient control. Summary of the Invention

[0015] To address the problems existing in the prior art, this invention provides a dynamic balancing system and method for tunnel face pressure based on multi-agent collaborative decision-making.

[0016] This invention is implemented as follows: a dynamic pressure balancing system for the tunnel face based on multi-agent collaborative decision-making, comprising:

[0017] The data acquisition module is used to collect real-time sensor data and ground settlement data from the tunnel boring machine.

[0018] The multi-agent decision-making central module includes a thrust control agent, a cutterhead control agent, a grouting support agent, and a global coordination agent. Each agent interacts with data and coordinates objectives through an attention mechanism. The global optimization objective of "tunneling efficiency - energy consumption - axis control - surface settlement" is decomposed into local sub-objectives for each agent through a value function decomposition algorithm. The thrust control agent is responsible for dynamic matching of face pressure, the cutterhead control agent is responsible for adaptive adjustment of cutting parameters, the grouting support agent is responsible for synchronous optimization of grouting volume and pressure, and the global coordination agent uses the reinforcement learning PPO algorithm to resolve decision conflicts and dynamically allocate objective weights among agents, updating the global decision strategy every 100ms.

[0019] The stability discrimination module is used to analyze the stress state and deformation of the working face based on multi-parameter data, using fuzzy logic algorithms or formation discrimination models, and to determine whether the working face is in a stable state.

[0020] The dynamic adjustment module is used to automatically adjust the tunneling speed and face pressure based on the output of the stability judgment module.

[0021] The special working condition processing module is used to automatically switch control strategies to smoothly transition the pressure at the working face when traversing complex strata or passing close to existing tunnels.

[0022] The display module is used to display the acquisition and judgment results in a visual manner.

[0023] Furthermore, the stability discrimination module:

[0024] (1) Construction of fuzzy theory model:

[0025] Determine the input variables: Based on engineering experience and experimental data, select the key factors affecting the stability of the tunnel face as input variables. Geological conditions (G), tunneling speed (V), tunnel face pressure (P), and support strength (S) are all continuous variables and need to be mapped to fuzzy sets through membership functions.

[0026] Fuzzy sets and membership functions:

[0027] Triangle membership function

[0028]

[0029] Parameter meanings: a (left endpoint), b (peak value), c (right endpoint); "good" geological conditions can be defined as a=80, b=90, c=100 (assuming the geological score range is 0-100).

[0030] Establish fuzzy sets: Define fuzzy sets for each input variable, including "good", "average", and "poor", and determine the membership function for each set;

[0031] Constructing a fuzzy rule base: Based on expert knowledge and experimental data, a fuzzy rule base is constructed to describe the relationship between input variables and the stability of the working face;

[0032] Fuzzy inference and defuzzification: Using the fuzzy inference mechanism, the fuzzy output of the stability of the tunnel face is inferred based on the fuzzy values ​​of the input variables, and then the fuzzy output is converted into specific values ​​or levels through the defuzzification method;

[0033] (2) Real-time formation discrimination function:

[0034] Data collection and preprocessing: Collect a large amount of stratigraphic data and working face stability cases, and perform data cleaning and feature extraction;

[0035] Function construction and training: A real-time stratigraphic discrimination function is constructed using machine learning algorithms (including support vector machines and neural networks), and the function parameters are optimized through training data;

[0036] Real-time discrimination and output: During construction, the current stratum data is input in real time, the trained discrimination function is used to quickly calculate the stability of the tunnel face, and the discrimination result is output.

[0037] (3) Computer software implementation:

[0038] Develop specialized software: Use programming languages ​​(including Python and C++) to develop specialized software to achieve rapid calculation of fuzzy theoretical models or real-time stratigraphic discriminant functions;

[0039] Integrated data interface: Integrates data interfaces with tunneling equipment and monitoring systems to acquire construction data in real time;

[0040] Visual output: The judgment results are visualized in the form of graphs and reports, which makes it easy for engineers to quickly understand the stability of the working face.

[0041] Furthermore, the dynamic adjustment module:

[0042] 1) High-precision actuator design:

[0043] Hydraulic control technology: Advanced hydraulic control technology is adopted, and high-precision hydraulic cylinders and hydraulic valve actuators are designed to achieve precise adjustment of tunneling speed and face pressure;

[0044] Electronic control technology: Combining electronic control technology, including servo motors and stepper motors, to design high-precision electric actuators, improving adjustment accuracy and response speed;

[0045] 2) Application of intelligent control algorithms:

[0046] PID control algorithm: The PID control algorithm is used to perform closed-loop control of tunneling speed and face pressure, and precise adjustment is achieved by adjusting the proportional, integral and derivative coefficients;

[0047] Other intelligent control algorithms include fuzzy control and neural network control. Based on the nonlinear and time-varying characteristics of the system, appropriate intelligent control algorithms are selected to improve control performance.

[0048] 3) Establishment of a feedback mechanism:

[0049] Real-time monitoring: The tunneling speed and face pressure parameters are monitored in real time through sensors;

[0050] Effect evaluation: Compare the monitoring data with the set values ​​to evaluate the effect of the adjustment;

[0051] Strategy optimization: Based on the evaluation results, further optimize and adjust the strategy, including adjusting the control algorithm parameters and modifying the fuzzy rule base.

[0052] Furthermore, the special working condition processing module:

[0053] 1) Special working condition analysis:

[0054] Experimental study: Analyze the fluctuation pattern of working face pressure under special working conditions through laboratory experiments or field tests;

[0055] Numerical simulation: Using numerical simulation software (including finite element analysis software) to simulate the stability of the tunnel face under special working conditions and verify the accuracy of the experimental results;

[0056] 2) Establishment of control model for special operating conditions:

[0057] Model Construction: Based on the analysis results of special working conditions, a special working condition control model is constructed to describe the relationship between face pressure and tunneling speed and support strength parameters under different working conditions.

[0058] Parameter adjustment: Based on the model output results, the control parameters are automatically adjusted, including increasing the support strength and reducing the tunneling speed, to cope with special working conditions;

[0059] 3) Real-time monitoring and activation control strategy:

[0060] Real-time monitoring: Real-time monitoring of changes in working conditions during construction, including sudden changes in geological conditions and equipment failures, through sensors;

[0061] Strategy activation: When a special operating condition is detected, the corresponding control strategy is activated in a timely manner, including switching to the special operating condition control mode and adjusting the control parameters;

[0062] Effectiveness verification: After implementing the control strategy, the stability changes of the tunnel face were monitored in real time to verify the effectiveness of the control strategy.

[0063] Another objective of this invention is to provide a method for intelligent control of shield tunnel face pressure based on multiple parameters, comprising:

[0064] Step 1: The data acquisition module is responsible for collecting real-time sensor data from the tunnel boring machine and surface settlement data.

[0065] Step 2: The main control module uses the stability discrimination module to make real-time discrimination of the working face stability based on fuzzy theory or real-time formation discrimination function; based on the collected multi-parameter data, it uses fuzzy logic algorithm or formation discrimination model to analyze the stress state and deformation of the working face and determine whether the working face is in a stable state.

[0066] Step 3: The dynamic adjustment module automatically and in real time adjusts the tunneling speed and face pressure based on the results of the stability judgment module.

[0067] Step 4: Through the special working condition processing module, corresponding control strategies are formulated for special application scenarios such as crossing complex composite strata (soft on top and hard on the bottom) interface areas, passing under existing operating tunnels or important pipelines at very close distances, to ensure a smooth transition of working face pressure and strict settlement control.

[0068] Step 5: Display the collected data information on the monitor using the display module.

[0069] Another object of the present invention is to provide a computer device, the computer device including a memory and a processor, the memory storing a computer program, the computer program being executed by the processor causing the processor to perform the steps of the intelligent control method for pressure at the face of a shield tunnel based on multiple parameters.

[0070] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the intelligent control method for pressure at the face of a shield tunnel based on multiple parameters.

[0071] Another objective of this invention is to provide an information data processing terminal for implementing the dynamic balancing system for face pressure based on multi-agent collaborative decision-making.

[0072] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows:

[0073] The core advantage of this invention lies in breaking the limitations of the traditional shield tunneling control system where the "thrust-cutterhead-grouting" system operates independently through a multi-agent collaborative mechanism. This transforms the dispersed execution units into a collaborative decision-making body with a shared goal, ultimately achieving an optimized balance of the global objectives of "tunneling efficiency, energy consumption, axis control, and surface settlement".

[0074] Real-time adaptive control: This system can automatically and in real time adjust the tunneling speed and face pressure based on real-time monitoring data and face stability judgment results, realizing "real-time adaptive" control, effectively coping with complex and ever-changing geological conditions, and improving the safety and stability of construction.

[0075] Effective handling of special working conditions: For special application scenarios such as traversing complex and composite strata, passing under existing buildings at extremely close distances, and abnormal tunneling states such as shield start-up, shutdown, and speed change, this system can formulate reasonable control strategies, smoothly transition the pressure at the tunnel face, strictly control settlement, and ensure construction safety and the safety of the surrounding environment.

[0076] Improving construction efficiency and quality: Precise parameter control reduces construction interruptions and rework caused by unreasonable parameters, thereby improving construction efficiency. Simultaneously, stable face pressure helps ensure tunnel excavation quality, reduces ground deformation and surface settlement, and enhances the overall stability of the tunnel.

[0077] Does the technical solution of this invention solve a technical problem that people have long desired to solve but have never been able to successfully solve? This system can automatically and in real time adjust the tunneling speed and face pressure based on real-time monitoring data and the results of face stability judgment, so as to achieve "real-time adaptive" control, effectively cope with complex and ever-changing geological conditions, and improve the safety and stability of construction. Attached Figure Description

[0078] Figure 1 This is a block diagram of a dynamic pressure balancing system for the tunnel face based on multi-agent collaborative decision-making, provided in an embodiment of the present invention.

[0079] Figure 2 This is a flowchart of the stability discrimination module method provided in the embodiment of the present invention.

[0080] Figure 3 This is a flowchart of the dynamic adjustment module method provided in an embodiment of the present invention.

[0081] Figure 4 This is a flowchart of an intelligent control method for shield tunnel face pressure based on multiple parameters, provided in an embodiment of the present invention.

[0082] Figure 1 The components are: 1. Data acquisition module; 2. Main control module; 3. Stability judgment module; 4. Dynamic adjustment module; 5. Special working condition handling module; 6. Display module. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0084] like Figure 1 As shown, an embodiment of the present invention provides a dynamic pressure balancing system for the tunnel face based on multi-agent cooperative decision-making, comprising:

[0085] Data acquisition module 1, main control module 2, stability judgment module 3, dynamic adjustment module 4, special working condition handling module 5, display module 6;

[0086] Data acquisition module 1, connected to main control module 2, is responsible for real-time acquisition of shield machine sensor data and ground settlement data;

[0087] The main control module 2 is connected to the data acquisition module 1, stability judgment module 3, dynamic adjustment module 4, special working condition processing module 5, and display module 6, and is used to control the normal operation of each module.

[0088] The stability discrimination module 3, connected to the main control module 2, is used to make real-time discrimination of the stability of the working face based on fuzzy theory or real-time formation discrimination function; based on the collected multi-parameter data, it uses fuzzy logic algorithm or formation discrimination model to analyze the stress state and deformation of the working face and determine whether the working face is in a stable state.

[0089] The dynamic adjustment module 4, connected to the main control module 2, is used to automatically and in real time adjust the tunneling speed and face pressure based on the results of the stability judgment module.

[0090] Special working condition processing module 5, connected to main control module 2, is used to formulate corresponding control strategies for special application scenarios such as crossing complex composite strata (soft on top and hard on the bottom), passing under existing operating tunnels or important pipelines at very close distances, to ensure a smooth transition of working face pressure and strict settlement control.

[0091] Display module 6 is connected to main control module 2 and is used to display the collected data information on the display.

[0092] Example 1: System Overall Architecture and Hardware Integration

[0093] This system uses an industrial computer as its core control unit, collecting multi-source sensor signals from the tunnel boring machine's (TBM) front end, including face pressure, soil chamber pressure, propulsion cylinder thrust, cutterhead torque, tunneling speed, and real-time data from the surface settlement monitoring instrument. The sensor signals are aggregated to the main control module via fieldbus (CAN or EtherCAT), and data standardization and time synchronization are performed through an A / D conversion interface. The main control module communicates with the ground monitoring center via Ethernet, enabling data upload and remote control, providing a highly timely foundation for subsequent judgment and adjustment algorithms.

[0094] Example 2: Implementation of Fuzzy Theory Stability Determination Algorithm

[0095] In the stability assessment module, the system selects four key parameters as input variables: geological conditions, tunneling speed, face pressure, and support strength. Each variable is fuzzified using a triangular membership function. For example, the geological condition variable is divided into three sets: "good," "average," and "poor," with the corresponding parameter ranges determined through on-site calibration. The system performs inference based on a fuzzy rule base. When the geological conditions are poor and the pressure is low, and the support strength is insufficient, the stability level of the face is determined to decrease. The centroid method is used for defuzzification, outputting a continuous stability index value, thus achieving the mapping from multidimensional data to quantitative results.

[0096] Example 3: Construction and Software Implementation of Real-Time Formation Discriminant Function

[0097] For different geological formations, the system pre-collects multiple sets of historical shield tunneling construction data and uses a neural network algorithm to train a geological formation discrimination function. Input variables include earth pressure, torque fluctuation, advance rate, and moisture content of the slag sample, and the output is the tunnel face stability level. A data training framework is built using Python, and the Adam optimization algorithm is used to iteratively update parameters. This is deployed in a real-time module written in C++ to ensure computational speed. The discrimination results are refreshed every second and called by the main control module to guide the execution of the dynamic adjustment module.

[0098] Example 4: Dynamic adjustment module and actuator linkage

[0099] The dynamic adjustment module connects to the hydraulic control unit via a CAN bus. The system sets the target pressure range at the tunnel face. When the real-time monitored value deviates from the set value, the PID controller automatically calculates the correction amount and adjusts the opening of the hydraulic valves in the propulsion cylinders to achieve stable pressure control. The electronic control actuator is responsible for driving the servo motor to adjust the tunneling speed, forming a speed-pressure dual closed-loop control system. The system also has a self-learning function, automatically correcting the PID parameters based on historical adjustment results to improve response speed and steady-state accuracy.

[0100] Example 5: Special Operating Condition Handling and Strategy Switching

[0101] When the system detects a sudden change in ground parameters or the tunnel boring machine (TBM) approaching an existing tunnel, the main control module automatically activates a special working condition control mode. In this mode, the special working condition processing module predicts the stress distribution trend at the tunnel face using a finite element model and dynamically adjusts the support pressure gain coefficient and the lower limit of the tunneling speed. For example, in a "soft upper layer and hard lower layer" composite stratum, the system prioritizes increasing the support strength and reducing the tunneling speed to minimize the risk of tunnel face instability. After the strategy is implemented, the system monitors the settlement rate in real time; if it stabilizes, it gradually reverts to the normal mode.

[0102] Example 6: Data Visualization and Human-Computer Interaction Interface

[0103] The display module employs a dual-channel output of an industrial touchscreen and a remote monitoring platform, showcasing real-time data on tunnel face pressure, torque, geological formation classification, and settlement trend curves. The interface integrates an alarm function; when the system detects stability indicators below a threshold, a red warning is issued and the system automatically logs the information. Engineers can query parameters, manually intervene, or switch modes through the interface, achieving full-process visual monitoring of the construction process and ensuring both tunnel boring safety and accurate control of tunnel face pressure.

[0104] like Figure 2 As shown, the stability discrimination module provided in this embodiment of the invention:

[0105] S101, Fuzzy Theory Model Construction:

[0106] Determine the input variables: Based on engineering experience and experimental data, select the key factors affecting the stability of the tunnel face as input variables. Geological conditions (G), tunneling speed (V), tunnel face pressure (P), and support strength (S) are all continuous variables and need to be mapped to fuzzy sets through membership functions.

[0107] Fuzzy sets and membership functions:

[0108] Triangle membership function

[0109]

[0110] Parameter meanings: a (left endpoint), b (peak value), c (right endpoint); "good" geological conditions can be defined as a=80, b=90, c=100 (assuming the geological score range is 0-100).

[0111] Establish fuzzy sets: Define fuzzy sets for each input variable, including "good", "average", and "poor", and determine the membership function for each set;

[0112] Constructing a fuzzy rule base: Based on expert knowledge and experimental data, a fuzzy rule base is constructed to describe the relationship between input variables and the stability of the working face;

[0113] Fuzzy inference and defuzzification: Using the fuzzy inference mechanism, the fuzzy output of the stability of the tunnel face is inferred based on the fuzzy values ​​of the input variables, and then the fuzzy output is converted into specific values ​​or levels through the defuzzification method;

[0114] S102, Real-time formation discrimination function:

[0115] Data collection and preprocessing: Collect a large amount of stratigraphic data and working face stability cases, and perform data cleaning and feature extraction;

[0116] Function construction and training: A real-time stratigraphic discrimination function is constructed using machine learning algorithms (including support vector machines and neural networks), and the function parameters are optimized through training data;

[0117] Real-time discrimination and output: During construction, the current stratum data is input in real time, the trained discrimination function is used to quickly calculate the stability of the tunnel face, and the discrimination result is output.

[0118] S103, Computer software implementation:

[0119] Develop specialized software: Use programming languages ​​(including Python and C++) to develop specialized software to achieve rapid calculation of fuzzy theoretical models or real-time stratigraphic discriminant functions;

[0120] Integrated data interface: Integrates data interfaces with tunneling equipment and monitoring systems to acquire construction data in real time;

[0121] Visual output: The judgment results are visualized in the form of graphs and reports, which makes it easy for engineers to quickly understand the stability of the working face.

[0122] like Figure 3 As shown, the dynamic adjustment module provided in this embodiment of the invention:

[0123] S201, High-precision actuator design:

[0124] Hydraulic control technology: Advanced hydraulic control technology is adopted, and high-precision hydraulic cylinders and hydraulic valve actuators are designed to achieve precise adjustment of tunneling speed and face pressure;

[0125] Electronic control technology: Combining electronic control technology, including servo motors and stepper motors, to design high-precision electric actuators, improving adjustment accuracy and response speed;

[0126] S202, Application of intelligent control algorithms:

[0127] PID control algorithm: The PID control algorithm is used to perform closed-loop control of tunneling speed and face pressure, and precise adjustment is achieved by adjusting the proportional, integral and derivative coefficients;

[0128] Other intelligent control algorithms include fuzzy control and neural network control. Based on the nonlinear and time-varying characteristics of the system, appropriate intelligent control algorithms are selected to improve control performance.

[0129] S203, Feedback mechanism established:

[0130] Real-time monitoring: The tunneling speed and face pressure parameters are monitored in real time through sensors;

[0131] Effect evaluation: Compare the monitoring data with the set values ​​to evaluate the effect of the adjustment;

[0132] Strategy optimization: Based on the evaluation results, further optimize and adjust the strategy, including adjusting the control algorithm parameters and modifying the fuzzy rule base.

[0133] The special working condition processing module provided in this embodiment of the invention:

[0134] 1) Special working condition analysis:

[0135] Experimental study: Analyze the fluctuation pattern of working face pressure under special working conditions through laboratory experiments or field tests;

[0136] Numerical simulation: Using numerical simulation software (including finite element analysis software) to simulate the stability of the tunnel face under special working conditions and verify the accuracy of the experimental results;

[0137] 2) Establishment of control model for special operating conditions:

[0138] Model Construction: Based on the analysis results of special working conditions, a special working condition control model is constructed to describe the relationship between face pressure and tunneling speed and support strength parameters under different working conditions.

[0139] Parameter adjustment: Based on the model output results, the control parameters are automatically adjusted, including increasing the support strength and reducing the tunneling speed, to cope with special working conditions;

[0140] 3) Real-time monitoring and activation control strategy:

[0141] Real-time monitoring: Real-time monitoring of changes in working conditions during construction, including sudden changes in geological conditions and equipment failures, through sensors;

[0142] Strategy activation: When a special operating condition is detected, the corresponding control strategy is activated in a timely manner, including switching to the special operating condition control mode and adjusting the control parameters;

[0143] Effectiveness verification: After implementing the control strategy, the stability changes of the tunnel face were monitored in real time to verify the effectiveness of the control strategy.

[0144] like Figure 4 As shown in the figure, an intelligent control method for shield tunnel face pressure based on multiple parameters provided by an embodiment of the present invention includes:

[0145] S301 is responsible for collecting real-time data from the tunnel boring machine's sensors and ground settlement data through its data acquisition module.

[0146] S302, the main control module uses the stability discrimination module to make real-time discrimination of the working face stability based on fuzzy theory or real-time formation discrimination function; based on the collected multi-parameter data, it uses fuzzy logic algorithm or formation discrimination model to analyze the stress state and deformation of the working face and determine whether the working face is in a stable state.

[0147] S303 automatically and in real time adjusts the tunneling speed and face pressure based on the results of the stability judgment module through the dynamic adjustment module;

[0148] S304, through a special working condition processing module, formulates corresponding control strategies for special application scenarios such as crossing complex composite strata (soft on top and hard on the bottom), passing under existing operating tunnels or important pipelines at extremely close distances, to ensure a smooth transition of working face pressure and strict settlement control.

[0149] S305 uses a display module to display the collected data information on a monitor.

[0150] Another object of the present invention is to provide a computer device, the computer device including a memory and a processor, the memory storing a computer program, the computer program being executed by the processor causing the processor to perform the steps of the intelligent control method for pressure at the face of a shield tunnel based on multiple parameters.

[0151] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the intelligent control method for pressure at the face of a shield tunnel based on multiple parameters.

[0152] Another objective of this invention is to provide an information data processing terminal for implementing the dynamic balancing system for face pressure based on multi-agent collaborative decision-making.

[0153] Specific implementation of the present invention:

[0154] Example 1: Traversing Complex Composite Strata

[0155] I. Project Overview (Continuing the original scenario, with additional parameter details)

[0156] A section of Metro Line 3 tunnel traverses a composite stratum of soft upper layer and hard lower layer. The upper layer is 3-5m thick silty clay (moisture content 35%-40%, cohesion 18kPa), and the lower layer is moderately weathered granite (uniaxial compressive strength 60-80MPa). The tunnel has an outer diameter of 6.0m and a burial depth of 15m. The section is located above an urban main road, and the allowable limit for surface settlement is ≤2.5mm.

[0157] II. System Application Process (Supplementing Key Parameters)

[0158] Data acquisition and working condition identification are achieved by capturing the difference in vibration dominant frequency (50-80Hz for soft soil and 180-220Hz for hard rock) using a cutterhead vibration sensor (model: Kistler 8793B500, sampling frequency 1000Hz). When the vibration dominant frequency is detected to jump from 75Hz to 200Hz within 3 seconds, it is determined to be the "soft upper and hard lower stratum interface", triggering the interface transition strategy.

[0159] Decision and execution parameters

[0160] Stability assessment: Input geological score 60 points, pressure deviation -0.2MPa, settlement rate 0.6mm / h, output stability level 3;

[0161] Multi-agent commands: The thrust control agent increases the pressure from 1.0 MPa to 1.3 MPa in steps (0.05 MPa increase for every 100 mm of tunneling); the cutterhead control agent reduces the rotational speed from 7 r / min to 5 r / min, and controls the torque at 1500-1700 kN·m; the grouting support agent increases the grouting volume from 2.2 m... 3 / Ring increased to 2.8m 3 / ring, pressure 0.5-0.7MPa.

[0162] Effect data prediction

[0163] Evaluation indicators Conventional regulation (forecast value) This system (predicted values) Optimization range Maximum settlement of the interface segment 4.2mm | 1.8mm Reduced by 57.1% Cutter head vibration amplitude ±0.8g ±0.3g Reduced by 62.5% Uneven force distribution in the propulsion cylinder 25% 8% Reduced by 68%

[0164] Example 2: Passing under an existing operating tunnel at extremely close proximity

[0165] I. Project Overview (Continuing the original scenario, with additional parameter details)

[0166] A highway shield tunnel needs to pass under an existing subway Line 2 tunnel with a clearance of only 2.8m. The existing tunnel was constructed using the shield tunneling method (outer diameter 5.4m), and the allowable settlement limit during operation is ≤0.5mm. The new tunnel has an outer diameter of 7.2m and a burial depth of 28m. The stratum in the section it will pass through is a layer of sand and gravel (particle size 50-150mm, medium-dense density).

[0167] II. System Application Process (Supplementing Key Parameters)

[0168] Data acquisition and working condition identification are achieved by monitoring the deformation of the existing tunnel structure through distributed fiber optic sensors (model: Corning SMF-28e+, spatial resolution 1m). When the newly built tunnel is 10m away from the existing tunnel, the system automatically determines it as an "underpass condition" and initiates a micro-disturbance control strategy.

[0169] Decision and execution parameters

[0170] Stability assessment: Input geological score 58 points, pressure deviation 0.15MPa, existing tunnel deformation 0.2mm, output stability level 2;

[0171] Multi-agent commands: The thrust control agent stabilizes the pressure at 1.6 MPa (fluctuation ±0.05 MPa) and reduces the tunneling speed to 15-18 mm / min (40% of the conventional speed); the cutterhead control agent adopts a low-torque mode (torque ≤1200 kN·m) and a rotation speed of 3-4 r / min; the grouting support agent uses synchronous grouting + secondary reinforcement grouting, with a synchronous grouting volume of 3.5 m. 3 / Ring (pressure 0.8-1.0MPa), secondary grouting volume 1.2m 3 / Ring (pressure 1.2-1.4MPa).

[0172]

[0173] Example 3: Crossing a Water-Rich Fracturing Zone

[0174] I. Project Overview

[0175] The shield tunnel section of Metro Line 5 in a certain city needs to traverse a water-rich fractured stratum approximately 180m long. This stratum is mainly moderately weathered sandstone with well-developed fissures (fissure rate 12%-15%), a water content of 25%-30%, and a groundwater level depth of 3.5m. There are risks of water inrush at the tunnel face, collapse, and excessive settlement. The tunnel has an outer diameter of 6.2m and a depth of 18m. There are no important buildings or structures within 50m of the traversing section, and the allowable surface settlement limit is ≤3mm.

[0176] II. System Application Process

[0177] Multi-source data acquisition and operating condition identification

[0178] The data acquisition module obtains key data in real time through the following sensors:

[0179] Face pressure sensor (model: Ashcroft 1005, range 0-2MPa, accuracy ±0.25% FS): sampling frequency 10Hz, real-time monitoring of face pressure fluctuations;

[0180] Formation water content sensor (model: TDR-300, measurement range 0-50%, accuracy ±1%): buried at 3 monitoring points around the cutterhead to capture sudden changes in water content;

[0181] Surface settlement monitoring instrument (model: Leica TS60, measurement accuracy ±0.1mm): one monitoring point is set up every 5m along the tunnel axis to record settlement data synchronously.

[0182] When the system detects that the moisture content suddenly increases from the normal 18% to 26% and the working face pressure fluctuates by ±0.35MPa within 10 seconds, the special working condition processing module automatically determines it as a "crossing a water-rich fracture zone" working condition using a CNN-LSTM hybrid model (with a recognition accuracy of 92.5%), triggering a dedicated collaborative strategy.

[0183] Stability assessment and decision output

[0184] The stability discrimination module is based on a fuzzy logic-reinforcement learning hybrid model. The input parameters are: geological condition score of 45 points (poor), face pressure deviation value of -0.3MPa, and surface settlement rate of 0.8mm / h. After fuzzification by the triangular membership function, combined with the multi-agent historical decision reward value (-6.2), the output face stability level is 4 (unstable).

[0185] Multi-agent decision-making central module response adjustment:

[0186] The global coordination agent increases the target weight of the grouting support agent from the conventional 25% to 40%, reduces the weight of the thrust control agent to 25%, and maintains the weight of the cutterhead control agent at 35%.

[0187] Output core control commands: Target face pressure 1.2MPa (fluctuation allowed ±0.1MPa), tunneling speed reduced to 20-25mm / min (50%-60% of conventional speed 40-45mm / min), cutterhead rotation speed 4-5r / min (50% of conventional speed 8-10r / min), dual-liquid grouting volume 3.5-4.0m. 3 / Ring (conventional grouting volume 2.0-2.5m) 3 (1.5 times the size of the ring), grouting pressure 0.6-0.8MPa.

[0188] Dynamic execution and effect monitoring

[0189] The dynamic execution control module stabilizes the face pressure at 1.18-1.22 MPa via an electro-hydraulic proportional valve (control accuracy ±0.01 MPa), while the servo motor precisely controls the tunneling speed at 22 mm / min and the cutterhead rotation speed at 4.2 r / min. The grouting support system switches to a dual-liquid grouting mode (cement grout to water glass volume ratio 1:0.8), with a grouting volume of 3.8 m³. 3 / ring, pressure 0.72MPa.

[0190] Continuous monitoring data shows that: the pressure fluctuation at the tunnel face is ≤ ±0.08MPa, the formation moisture content is stable at 26%-28%, and there is no water inrush; the maximum surface settlement is 2.1mm (located in the middle of the fracture zone), which is far below the allowable limit of 3mm; the tunnel convergence is 0.8mm, which meets the structural stability requirements.

[0191] III. Technical Effect Data Prediction

[0192]

[0193] Example 4: Extremely Cold Strata Scenario

[0194] I. Project Overview

[0195] A shield tunnel for an intercity railway in northern China needs to traverse a 250m long section of extremely cold permafrost. This stratum is mainly composed of silty clay, with a freezing depth of 8-12m. During construction, the ambient temperature ranges from -15℃ to -8℃, the natural moisture content of the permafrost is 20%-23%, and the unfrozen water content is 5%-8%. Risks include cutterhead adhesion, freezing of grouting pipelines, and frost heave instability at the tunnel face. The tunnel has an outer diameter of 8.8m and a burial depth of 22m. The surface of the traversed section is farmland, with a permissible settlement limit of ≤4mm.

[0196] II. System Application Process

[0197] Multi-source data acquisition and operating condition identification

[0198] The data acquisition module focuses on monitoring the following parameters:

[0199] Ambient temperature sensor (model: PT100, measurement range -50℃-100℃, accuracy ±0.2℃): installed in the cutterhead, grouting pipeline and soil chamber of the tunnel boring machine to record temperature changes in real time;

[0200] Tool turret torque sensor (model: RDP T40B, range 0-3000kN·m, accuracy ±0.5% FS): monitors changes in tool turret cutting resistance;

[0201] Grouting pipeline pressure sensor (model: HBM U9C, range 0-10MPa, accuracy ±0.1% FS): detects sudden pressure rises caused by pipeline blockage.

[0202] When the system detects that the temperature in the cutterhead area drops to -12℃, the cutterhead torque increases from the usual 1200kN·m to 1800kN·m (signs of adhesion), and the grouting pipeline outlet temperature is -9℃, the special working condition processing module automatically determines it as an "extremely cold formation" working condition and activates the antifreeze and anti-adhesion synergistic strategy.

[0203] Stability assessment and decision output

[0204] Input parameters for the stability discrimination module: geological condition score 55 (general), face pressure deviation value 0.2MPa, surface settlement rate 0.5mm / h. After fuzzification, combined with the reward value (-3.8), the output stability level is 3 (critical stability).

[0205] Multi-agent decision-making central module adjustment strategy:

[0206] The global coordination agent increases the target weight of the cutterhead control agent from 30% to 38%, the weight of the grouting support agent to 32% (focusing on pipeline antifreeze), and the weight of the thrust control agent to 30%.

[0207] Output core control commands: Increase cutterhead speed to 7-8 r / min (1.4 times the conventional speed of 5-6 r / min); cutterhead torque control upper limit 2000 kN·m; target face pressure 1.5 MPa (fluctuation ±0.15 MPa), using a stepped adjustment (pressure increased by 0.05 MPa for every 500 mm of tunneling); grouting fluid temperature controlled at 18-20℃ (achieved through pipeline heating device); grouting pressure 0.9-1.1 MPa; grouting volume 4.2-4.5 m³. 3 / Ring.

[0208] Dynamic execution and effect monitoring

[0209] The dynamic control module activates the cutterhead heating element (5kW) to maintain the cutterhead surface temperature between -5℃ and -3℃, preventing frozen soil from sticking. The servo motor stabilizes the cutterhead speed at 7.5 r / min and controls the torque between 1700-1850 kN·m. The grouting system uses an electric heating jacket (3kW) to raise the grout temperature to 19℃, with a grouting pressure of 1.0 MPa and a grouting volume of 4.3 m. 3 The pipeline showed no signs of freezing or blockage.

[0210] Continuous monitoring shows that the working face pressure is stable at 1.48-1.53 ​​MPa, the maximum surface settlement is 3.2 mm (below the 4 mm limit), the number of cutterhead adhesions has decreased from 6 times / day under normal control to 0 times / day, the grouting pipeline has good flow, there is no sudden pressure rise, and the tunnel excavation efficiency reaches 5.2 m / day (close to 90% of the efficiency of normal temperature strata).

[0211] III. Technical Effect Data Prediction

[0212]

[0213] I. Specific application areas or related products of this invention.

[0214] Urban subway tunnel construction

[0215] In complex geological conditions such as soft soil, gravel, and composite strata, this invention reduces surface subsidence and ensures the safety of surrounding buildings by real-time control of the working face pressure. For example, similar sealing technologies have been applied in projects such as Beijing Subway Line 12 and the Pearl River Delta Water Resources Project; this invention can further optimize the accuracy of pressure control.

[0216] Special scenarios: When passing under existing operating tunnels or important pipelines at extremely close distances (such as subway tunnels crossing underground utility tunnels), settlement control must be strictly controlled to the millimeter level.

[0217] Mountain railway and highway tunnel engineering

[0218] To address geological conditions such as high water pressure, fault fracture zones, and karst development, this invention dynamically adjusts tunneling speed and pressure parameters to prevent face instability. For example, the Jintang Subsea Tunnel of the Ningbo-Zhoushan Railway (maximum water pressure 0.85 MPa) needs to cope with complex seabed strata, and this invention can improve its construction safety.

[0219] Technology adaptation: Based on the intelligent construction practice of long high-speed railway shield tunnels in the Yangtze River Delta region, optimize advanced geological forecasting and parameter matching.

[0220] II. Evidence related to the technical effects obtained by the embodiments of the present invention.

[0221] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0222] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A system for dynamic balance of face pressure based on multi-agent collaborative decision making, characterized in that, The application relates to a shield tunneling machine control system, comprising: a data acquisition module for real-time acquisition of shield sensor data and ground settlement data; a multi-agent decision-making hub module containing a thrust control agent, a cutterhead control agent, a grouting support agent and a global coordination agent, each agent realizing information interaction and target coordination based on an attention mechanism; the multi-agent decision-making hub module decomposes the global optimization target of tunneling efficiency, energy consumption, axis control and ground settlement into multiple local sub-targets through a value function decomposition algorithm, wherein the thrust control agent is responsible for dynamic matching of the working face pressure, the cutterhead control agent is responsible for adaptive adjustment of the cutting parameters, the grouting support agent is responsible for collaborative optimization of the grouting amount and pressure, and the global coordination agent performs conflict resolution and weight distribution based on a reinforcement learning strategy update algorithm; a stability discrimination module for judging the stability state of the working face according to the stratum and tunneling parameters; a dynamic adjustment module for automatically adjusting the tunneling speed and working face pressure according to the discrimination result; a special working condition processing module for automatically switching the control strategy to maintain pressure balance when complex strata or existing tunnels are crossed; a display module for visual display of the collected data and discrimination results.

2. The system of claim 1, wherein, The stability discrimination module constructs input variables based on fuzzy theory, the input variables including geological conditions, tunneling speed, working face pressure and support strength, which are all mapped to fuzzy sets through membership functions to realize quantitative-qualitative conversion.

3. The system of claim 2, wherein, The membership function adopts a triangular function form, the parameters of which include a left end point value, a peak value and a right end point value, different fuzzy sets are set according to geological grade intervals, and the membership degree of the working face stability state is represented.

4. The system of claim 1, wherein, The stability discrimination module can construct a real-time stratum discrimination function based on a support vector machine or a neural network, obtain a classification boundary through training of stratum sample data, and calculate the working face stability in real time during tunneling.

5. The system of claim 1, wherein, The dynamic adjustment module comprises an actuator unit, a control algorithm unit and a feedback unit; the control algorithm unit adopts a proportional-integral-derivative control algorithm, a fuzzy control or a neural network control algorithm, and realizes closed-loop accurate adjustment of the pressure and speed by dynamically adjusting the proportional coefficient, the integral coefficient and the differential coefficient; the feedback unit monitors the adjustment result in real time and optimizes the control parameters.

6. A multi-parameter-based intelligent control method for shield tunnel face pressure, characterized in that, The application further discloses a shield tunneling machine control method, comprising the following steps: Step 1: collecting shield sensor data and ground settlement data; Step 2: calculating the working face stability grade based on a fuzzy logic algorithm or a stratum discrimination function; Step 3: automatically adjusting the tunneling speed and working face pressure according to the stability grade; Step 4: switching a special control strategy to maintain stable working face pressure when complex strata or abnormal working conditions are detected; Step 5: visualizing the real-time data, control parameters and stability results.

7. The method of claim 6, wherein, The fuzzy logic algorithm comprises three stages of fuzzification, rule reasoning and defuzzification, wherein the defuzzification adopts a gravity center method to convert the fuzzy output into a stability numerical grade for automatic control.

8. The method of claim 6, wherein, The control process optimizes the decision-making strategies of the agents through a reinforcement learning algorithm, updates the action probability distribution through a proximal policy optimization algorithm, and performs global policy update once every hundred milliseconds.

9. The method of claim 6, wherein, The special working condition control includes three stages of working condition analysis, model establishment and strategy switching; In the working condition analysis stage, the numerical simulation or experimental method is used to identify the trend of formation parameter change; In the model establishment stage, a nonlinear function relationship between the face pressure and the driving speed and support strength is constructed; In the strategy switching stage, the control parameters are adjusted in real time according to the function output to realize the smooth transition of pressure.

10. The method of claim 9, wherein, The nonlinear function relationship is constructed based on the finite element analysis results, describes the coupling law between the face pressure change and the driving parameters in the composite stratum with soft upper and hard lower, and is used to guide the stability optimization of the dynamic control process.

Images