A system for coordinated regulation of spoil conveying and soil chamber pressure in shield construction

Abstract

The present application relates to the technical field of regulation system, disclose a kind of shield construction slag conveying and soil bin pressure coordinated regulation system, including multi-section data acquisition module: for in shield soil bin, helical conveyor multiple interval positions along the axial direction and the place of soil outlet respectively gather operating parameter, the operating parameter includes: soil bin pressure and its time variation rate, the driving load parameter corresponding to each axial position of helical conveyor, energy consumption parameter in conveying process and the continuity parameter of soil removal;Along the characteristic construction module: for based on the operating parameter constructs the along the state characteristic of helical conveying channel.The present application can be through the construction multi-section operating parameter perception system and along the state characteristic analysis model, realize the fine description of helical conveying process, and establish the quantitative relationship between conveying capacity and soil bin pressure demand, to improve the response capability and control precision of complex working condition in shield construction process, guarantee construction safety and stable operation.

Description

Technical Field

[0001] This invention relates to the field of control system technology, specifically to a coordinated control system for the transport of excavated soil and the pressure of the soil chamber during shield tunneling construction. Background Technology

[0002] Shield tunneling is widely used in urban subway and underground utility tunnel projects. Especially in earth pressure balance (EPB) shield tunneling, the stable control of earth chamber pressure plays a crucial role in maintaining face stability, preventing surface subsidence, and ensuring construction safety. Simultaneously, the excavated soil must be continuously discharged via a screw conveyor, and its conveying efficiency and stability directly affect the dynamic balance of earth chamber pressure. Due to the complex and variable properties of the excavated soil and the influence of various factors during conveyance, a significant nonlinear coupling relationship exists between earth chamber pressure and excavated soil conveyance. Therefore, achieving coordinated control of these two factors has always been a critical technical challenge in shield tunneling construction control.

[0003] Existing technologies often employ experience-based or single-parameter control methods, such as controlling the amount of soil discharged by adjusting the screw conveyor's rotational speed, or maintaining the pressure in the soil silo by adjusting the propulsion speed and force. However, these methods typically rely on only a limited amount of monitoring data, lacking a comprehensive understanding of the internal conditions along the screw conveyor channel, and are unable to accurately reflect the load distribution, trends, and local anomalies of the excavated soil during transport. Furthermore, existing methods provide a rather coarse characterization of transport capacity, failing to comprehensively consider multi-dimensional characteristics such as load, energy consumption, and the continuity of soil discharge, resulting in inaccurate assessments of actual transport capacity. In terms of control strategies, soil silo pressure control and soil discharge control are often independent of each other, lacking a unified and coordinated adjustment mechanism. When stratum conditions or operating conditions change, a mismatch between soil discharge capacity and pressure maintenance requirements can easily occur, leading to risks such as soil silo pressure drop, gushing, or screw conveyor blockage.

[0004] Furthermore, existing control systems mostly employ fixed thresholds or simple feedback adjustment methods, lacking sufficient dynamic adaptability and making it difficult to perform phased regulation based on changes in operating status. This can easily lead to control lag or improper adjustment during transitional or abnormal operating conditions. Simultaneously, for abnormal sections in the conveying process, current technologies lack effective identification and propagation analysis methods, making it difficult to predict abnormal development trends and take targeted measures in a timely manner. Moreover, the lack of collaborative optimization and closed-loop adjustment mechanisms among multiple parameters, coupled with the relative separation of risk identification and control strategies, makes it difficult to dynamically adjust the control range and implement tiered safety control during risk development. Therefore, there is an urgent need for a technical solution capable of achieving multi-section status perception, accurate characterization of conveying capacity, and multi-parameter collaborative closed-loop regulation to improve the stability, safety, and intelligence level of the tunnel boring machine (TBM) construction process. Summary of the Invention

[0005] Based on the above description, the present invention provides a coordinated control system for the transport of excavated soil and the pressure of the soil chamber in shield tunneling construction. By constructing a multi-segment operation parameter sensing system and a process state characteristic analysis model, the system achieves a refined description of the screw conveying process and establishes a quantitative relationship between the conveying capacity and the pressure demand of the soil chamber. This improves the response capability and control accuracy of shield tunneling in response to complex working conditions, and ensures construction safety and stable operation.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: a shield tunneling excavation soil transportation and soil chamber pressure coordinated control system, comprising: a multi-segment data acquisition module: used to collect operating parameters at multiple intervals along the axial direction of the shield soil chamber, the screw conveyor and the soil outlet, respectively. The operating parameters include: soil chamber pressure and its time change rate, drive load parameters corresponding to each axial position of the screw conveyor, energy consumption parameters during the transportation process and soil discharge continuity parameters.

[0007] The process feature construction module is used to construct the process state features of the spiral conveyor channel based on the operating parameters. The process state features include at least: load distribution parameters determined according to the driving load parameters, load gradient parameters calculated based on the load distribution parameters, load evolution rate parameters calculated based on the time change of the load gradient parameters, and conveying stability parameters calculated based on the energy consumption parameters and the soil discharge continuity parameters.

[0008] Equivalent conveying capacity calculation module: used to calculate the equivalent conveying capacity parameters characterizing the conveying capacity of the screw conveyor system based on the load distribution parameters, conveying stability parameters, and energy consumption parameters;

[0009] The conveying status and capacity determination module is used to determine the conveying status category and its development trend of slag and soil based on the load gradient parameter and load evolution rate parameter, determine the soil silo pressure maintenance requirement parameter based on the soil silo pressure change rate, and determine the degree of mismatch between conveying capacity and pressure maintenance requirement based on the relationship between the equivalent conveying capacity parameter and the soil silo pressure maintenance requirement parameter.

[0010] Dynamic control window generation module: used to generate a multi-parameter dynamic collaborative control window based on the soil transport status category, development trend and mismatch degree. The control window includes at least: allowable range of soil discharge capacity, allowable range of tunneling advance and soil improvement adjustment range.

[0011] Deviation determination module: used to determine the system operation deviation and its degree based on the relationship between the current operating parameters and the multi-parameter dynamic collaborative control window;

[0012] Closed-loop reconfiguration control module: used to perform the following operations when there is an operational deviation: based on the load gradient parameters and load evolution rate parameters, determine the abnormal section and its influence range; based on the influence range, establish the coupling relationship between each section in the spiral conveying channel; based on the degree of mismatch, coordinately adjust the soil discharge capacity parameters, soil chamber pressure maintenance parameters and tunneling advancement parameters; and gradually adjust the soil conveying state according to the preset adjustment sequence so that the system evolves from an abnormal state to a stable conveying state.

[0013] Steady-state maintenance module: used to maintain the coordinated and stable operation of the slag conveying and soil pressure when the operating parameters are within the multi-parameter dynamic coordinated control window.

[0014] Compared with the prior art, the technical solution of this application has the following beneficial technical effects:

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] 1. Through multi-segment data acquisition and feature construction along the route, a refined perception of the internal state of the screw conveyor channel is achieved, which can accurately reflect the load distribution and its changing characteristics during the transportation of slag and soil, and improve the accuracy of state identification.

[0017] 2. By constructing a capacity-resistance coupling model and introducing energy consumption and soil discharge continuity parameters, a quantitative characterization of the slag transport capacity was achieved, overcoming the problem of coarse transport capacity assessment in traditional methods.

[0018] 3. By establishing a matching relationship between conveying capacity and earth chamber pressure demand, and constructing a multi-parameter dynamic collaborative control window, the coordinated regulation of earth removal capacity, propulsion parameters and improvement parameters was realized, thereby improving the overall control effect of the system.

[0019] 4. By identifying abnormal sections and analyzing their propagation, the system can predict abnormalities in advance. Combined with a closed-loop reconfiguration control mechanism, the system can quickly adjust and restore stable operation under abnormal conditions. Attached Figure Description

[0020] Figure 1 This is a block diagram of the overall structure of a shield tunneling excavation soil transportation and soil chamber pressure coordinated control system according to the present invention;

[0021] Figure 2 This is a schematic diagram of the multi-section monitoring layout of the shield tunnel soil chamber, screw conveyor, and soil outlet in this invention;

[0022] Figure 3 This is a schematic diagram illustrating the identification and propagation of abnormal sections in the spiral conveying channel in this invention;

[0023] Figure 4This is a schematic diagram illustrating the calculation of the equivalent conveying capacity in this invention;

[0024] Figure 5 This is a schematic diagram illustrating the generation of the multi-parameter dynamic collaborative control window in this invention;

[0025] Figure 6 This is a flowchart of the closed-loop reconfiguration control process in this invention;

[0026] Figure 7 This is a flowchart of the risk protection and control process in this invention;

[0027] Figure 8 This is a flowchart of a method for coordinated control of excavated soil transportation and soil pressure in shield tunneling, according to the present invention. Detailed Implementation

[0028] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings, which illustrate embodiments of the present application. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of this application will be thorough and complete.

[0029] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0030] refer to Figure 1 and Figure 2 Example 1: Standard operating condition coordinated control applicable to clay-silty clay composite strata in urban subways

[0031] Project conditions and overall equipment configuration: The tunnel boring machine has an excavation diameter of 6.48m and an outer diameter of 6.28m. The cutterhead power is 900kW. The total propulsion system uses 16 sets of propulsion cylinders, with a rated thrust of 2500kN per set and a total rated thrust of 40000kN. The screw conveyor adopts a single-screw variable frequency drive structure. The screw blades have an outer diameter of 800mm, a pitch of 640mm, a blade thickness of 28mm, a central shaft diameter of 273mm, an effective conveying length of 9.6m, a rated maximum speed of 22rpm, a common operating speed of 6-16rpm, a drive motor rated power of 160kW, a reducer transmission ratio of 1:49, and an electro-hydraulic proportional control gate for the excavation. The stratum consists of plastic to stiff plastic clay. The soil is predominantly silty clay with a natural moisture content of 19%–27%, a liquid limit of 32%–41%, a plasticity index of 12–19, a permeability coefficient of 1.5×10⁻⁷–4.8×10⁻⁶ m / s, a cover thickness of 12.6m–18.4m, and a groundwater depth of approximately 2.8m. The designed target pressure control range for the soil chamber is 0.18–0.24 MPa. The conventional tunneling speed is controlled at 25–45 mm / min, and the width of a single-ring segment is 1.5m. The system uses a Siemens S7-1500H redundant PLC as the underlying controller, an Advantech IPC-610L industrial control host, an Advantech UNO-2484G edge computing terminal, and an ABB frequency converter. The ACS880-01-206A-3 system utilizes a WIKA S-20 series 0-1MPa diffused silicon pressure transmitter for the soil chamber pressure sensor, an HBM T40B torque sensor for segmented torque measurement of the screw conveyor, a Siemens SENTRON PAC3200 for power monitoring, a Mettler Toledo IND780 with an ICS429 weighing frame for the outlet belt scale, a Graco Husky 1050 pneumatic diaphragm pump for foam injection, a Schwing BP series screw pump for bentonite grouting, and a ProMinent gamma / X metering pump for polymer injection. All sensors are connected to the edge processing unit via PROFINET and Modbus TCP, with a sampling period of 200ms, a feature calculation period of 1s, a control update period of 2s, and a risk assessment period of 5s. This engineering background and system concept align with the overall technical direction outlined in your provided document: "multi-segment operating parameter sensing, along-process feature analysis, equivalent conveying capacity calculation, dynamic collaborative control, and risk control."

[0032] Multi-segment data acquisition module: The multi-segment data acquisition module is arranged around a three-segment path: soil chamber—screw conveyor—soil outlet. Inside the soil chamber, four pressure measuring points P1–P4 are arranged circumferentially, located at 15° above, 20° below, 90° to the left, and 90° to the right of the cutterhead center, respectively, with a sampling range of 0–1 MPa and an accuracy class of 0.25%. Simultaneously, two pressure change rate calculation channels are set up, and the edge processing unit performs sliding differential calculations on P1–P4 to obtain the instantaneous pressure change rate dP / dt. The screw conveyor is divided into five monitoring segments A, B, C, D, and E along the axial direction, with segment A near the soil chamber inlet and segment E near the soil outlet. Each segment is equipped with drive load acquisition points, including speed feedback, torque estimation, current, voltage, vibration acceleration, and bearing temperature. To improve the segment sensing accuracy, additional cylinder wall strain gauge groups (model HBM LY11-6 / 120) are installed in segments A and C to collect local shell load responses. Energy consumption parameters are estimated using motor input power, reducer output power, power consumption per unit time, and energy consumption per unit volume of soil transported. Discharge continuity parameters are composed of the instantaneous mass flow rate of the belt scale at the discharge port, time interval pulsation coefficient, and image recognition continuity flag values. The image recognition unit uses a Hikvision MV-CH120-10GC industrial camera at 30fps to identify whether the discharged material is continuous, whether large agglomerates exist, and whether there are periodic flow interruptions. All acquired parameters are input into the upper-level model after Kalman filtering and 3σ outlier removal. To meet the acquisition requirements of "soil chamber pressure and its rate of change over time, drive load parameters at each axial position, energy consumption parameters during transport, and discharge continuity parameters," the system uses a unified clock synchronization in its hardware layout and data link, with a timestamp accuracy better than 10ms. The time base of all sensors is unified to the PTP time synchronization module.

[0033] The data acquisition method for each section is no longer limited to the traditional extensive monitoring of a single soil chamber pressure or a single spiral current. Instead, a continuous parameter sensing network covering the soil chamber, the conveying process, and the discharge end has been established. On the one hand, it can more accurately identify subdivided phenomena such as congestion at the spiral conveyor inlet, local adhesion in the middle section of the cylinder, and intermittent flow interruption at the discharge end. On the other hand, changes in soil chamber pressure, torque, energy consumption fluctuations, and the continuity of soil discharge can be analyzed synchronously over time. This provides a reliable data foundation for subsequent load distribution modeling, abnormal section identification, and capacity-pressure matching calculation, significantly improving the pertinence and predictability of the control strategy.

[0034] The feature construction module along the path is deployed between the edge processing unit and the central control unit, adopting a two-level structure of "edge pre-extraction + central fine calculation". The edge processing unit first normalizes the original drive load parameters of the five sections A to E, forming the original load vector for each section: Li(t)=[Ii(t),Ti(t),ni(t),ai(t),θi(t)]; where Ii is the drive current, Ti is the estimated or measured torque, ni is the actual rotational speed of the screw, ai is the vibration intensity, and θi is the bearing temperature. The central control unit further calculates the comprehensive load value of the section according to the weighted mapping relationship:

[0035] L'i(t) = 0.32·Ii* + 0.38·Ti + 0.12·(1-ni^) + 0.10·ai* + 0.08·\thetai*, where variables marked with "*" represent standardized dimensional values. The axial center positions of segments A to E are 0.9m, 2.6m, 4.4m, 6.8m, and 8.9m, respectively. The load gradient parameter Gi = (L'(i+1)-L'i) / (x*(i+1)-xi) is calculated based on the comprehensive load values ​​of adjacent segments. In the time dimension, linear regression is performed on each gradient parameter within a 10s sliding window to obtain the load evolution rate Vi = dGi / dt. The transport stability parameter St is composed of the unit volume energy consumption Ev, the soil discharge continuity coefficient Cc, and the soil discharge discontinuity frequency fb, taken as:

[0036] St=0.45·(1-Ev*)+0.35·Cc+0.20·(1-fb^); where Ev=power consumption during the cycle / soil discharge volume during the cycle, soil discharge continuity coefficient Cc is defined as the ratio of continuous soil discharge time to total monitoring time, and soil discharge interruption frequency fb is defined as the number of interruption events per unit minute. Through the above calculations, the system can form the along-path state characteristics of the spiral conveying channel, including load distribution parameters, load gradient parameters, load evolution rate parameters, and conveying stability parameters. Under normal operating conditions, the typical data obtained from monitoring in Example 1 are as follows: the comprehensive load values ​​of sections A to E are 0.42, 0.47, 0.51, 0.49, and 0.44, respectively; the main gradient parameters are GAB=0.029, GBC=0.022, GCD=-0.008, and GDE=-0.021; the absolute value of the evolution rate of each gradient within a 10s window is usually less than 0.006 / s; the energy consumption per unit volume Ev is 0.82 to 1.16 kWh / m^3; the soil discharge continuity coefficient Cc is above 0.91, and the soil discharge interruption frequency is less than 0.4 times / min. This state corresponds to the basic characteristic of "stable conveying". By converting the original collected data into a unified state characteristic along the conveyor, problems such as "current rise but cause unknown" and "unstable soil discharge but location unknown" are transformed into calculable problems such as "which section has abnormal load, in which direction does the gradient change abruptly, and whether the abnormality is accelerating or decaying". This not only improves the interpretability of the internal state of the screw conveyor, but also makes subsequent control decisions no longer rely on single-point experience, but are based on quantifiable, comparable and predictable state parameters.

[0037] Implementation of the Equivalent Conveying Capacity Calculation Module: The core task of the equivalent conveying capacity calculation module is to calculate the equivalent conveying capacity parameter Qeq, which characterizes the overall soil discharge capacity, based on the load along the conveying path, conveying stability, and energy consumption information. The spiral conveying channel is divided into five axial segments. For each segment, the following parameters are calculated: segment load characteristic parameter, segment load gradient parameter, segment load non-uniformity coefficient, segment load evolution rate parameter, unit volume conveying energy consumption parameter, soil discharge continuity coefficient, and soil discharge discontinuity frequency parameter. The segment load non-uniformity coefficient Ui is defined as: Ui = σi / (μi + ε); where σi is the standard deviation of the segment load within a preset time window, μi is the average value of the segment load, and ε is set to 0.001 to prevent the denominator from being zero. The segment load evolution rate parameter Vi is the average value of the first derivative of the load gradient change within a 15-second window. The conveying resistance parameter Ri is calculated using the following formula: Ri = 0.40·|Gi*| + 0.30·Ui + 0.30·|Vi^|; the conveying capacity parameter Ci is calculated using the following formula: Ci = 0.45·(1-Evi*) + 0.35·Cci + 0.20·(1-fbi^); where Evi is the equivalent unit volume conveying energy consumption of each section, Cci is the section's soil discharge continuity coefficient, and fbi is the section's soil discharge discontinuity frequency. Considering the different impacts of different axial sections on the overall conveying capacity, this embodiment sets the initial weight coefficient W = [0.28, 0.24, 0.20, 0.16, 0.12], meaning that the closer a section is to the soil silo inlet, the greater its weight. When the load gradient of a certain section continuously increases in a positive direction and the rate of change is greater than 0.008 / s, the weight of that section is increased by 5% to 15%, the weight of the adjacent upstream section is increased by 3% to 8%, and the weights of the remaining sections are reduced in a normalized manner. After weighted coupling of each section, the combined capacity-resistance quantity is obtained: H=Σ(Wi·(Ci-Ri));

[0038] The equivalent transport capacity parameters are then calculated using a preset capacity-resistance mapping relationship: Qeq=Qref·[1+αH-βH^2]; where Qref is the baseline soil discharge capacity, taken as 145 m³ / h in Example 1; α is 0.62, and β is 0.18. After on-site calibration, Qeq≥138 m³ / h is considered "sufficient capacity", 122~138 m³ / h is considered "basically matched capacity", 108~122 m³ / h is considered "tight capacity", and less than 108 m³ / h is considered "insufficient capacity". To ensure model robustness, the system updates α and β in the capacity-resistance mapping relationship every 50 cycles, using the recursive least squares method, and uses the actual soil discharge volume and theoretical soil pressure of the current period to maintain consistency constraints for correction.

[0039] By defining the equivalent conveying capacity through a capacity-resistance coupling method, the one-sidedness of estimating the soil discharge capacity solely based on the screw speed or current in traditional methods is avoided. It can simultaneously reflect two dimensions: "how much effort is required for conveying" and "how stable is the conveying." It can amplify the impact of key congested sections on the overall capacity assessment through a section weight adaptive adjustment mechanism. It can also correct model parameters online through historical loop sections, making the capacity assessment more consistent with the actual on-site working conditions.

[0040] The conveying status and capacity determination module uses the equivalent conveying capacity parameter Qeq, the load gradient parameter Gi, the load evolution rate parameter Vi, and the soil pressure change rate dP / dt as core inputs. The system classifies the soil conveying status into five categories: stable conveying, local obstruction, overall eccentric loading, intermittent conveying, and blockage development. The specific judgment rules are as follows: When |max(Gi)|≤0.035, |Vi|≤0.006 / s, Cc≥0.90 and fb≤0.5 times / min, it is judged as stable transport; when the absolute value of the gradient parameter of a single section is greater than 0.045 and increases in the same direction for three consecutive sampling periods, while the adjacent sections are still within the normal range, it is judged as local blockage; when the gradient of multiple consecutive sections rises as a whole, and the average value of the load non-uniformity coefficient is greater than 0.085, it is judged as overall unbalanced load; when the soil discharge continuity coefficient is lower than 0.85 and the soil discharge interruption frequency is greater than 1.2 times / min, it is judged as intermittent transport; when the Vi of a candidate abnormal section is higher than 0.012 / s for 10 consecutive seconds and the abnormal influence range continues to expand, it is judged as blockage development.

[0041] The soil pressure maintenance requirement parameter Pd is determined by the current soil pressure Pc, target pressure Pset, pressure change rate dP / dt, and propulsion speed vj, using the following formula: Pd = k1(Pset - Pc) + k2·|dP / dt| + k3·vj; where k1 = 0.55, k2 = 0.30, and k3 = 0.15. Then, the mismatch degree between transport capacity and pressure demand Dm is determined based on the relationship between Qeq and Pd: Dm = (Qeq - Qdem) / Qdem; where Qdem is the target soil discharge capacity converted according to the maintenance pressure requirement, Qdem = η1·Pd + η2·vj + η3·Fp, Fp is the converted term for total propulsion thrust, η1 = 280, η2 = 1.8, and η3 = 0.0009. If Dm ≥ 0.10, it indicates a surplus of capacity; -0.08 ≤ Dm < 0.10 indicates a basic match; -0.20 ≤ Dm < -0.08 indicates a slight mismatch; and Dm < -0.20 indicates a severe mismatch. This result is directly used by the subsequent dynamic control window generation module.

[0042] During a certain tunneling cycle, the target pressure of the earth chamber Pset = 0.22 MPa, the actual average pressure Pc = 0.208 MPa, the pressure change rate dP / dt = -0.006 MPa / s, the advance speed vj = 36 mm / min, and the total thrust Fp = 23600 kN. The calculated Pd = 0.55 × 0.012 + 0.30 × 0.006 + 0.15 × 36 / 1000 ≈ 0.0138. If calculated according to the project calibration coefficient, Qdem = 280 × 0.0138 + 1.8 × 36 + 0.0009 × 23600 ≈ 90.19 m³ / h. At this time, Qeq = 146.8 m³ / h, then Dm = (146.8 - 90.19) / 90.19 ≈ 0.628, which falls under the category of "capacity surplus". In this state, the system will not blindly increase the discharge of soil, but will judge whether to maintain the current speed or slightly reduce the speed based on the change rate of soil pressure in the soil chamber, so as to avoid excessive discharge leading to pressure drop.

[0043] The system is configured with a segmented coupling control analysis unit. First, based on the 9.6m long structure of the screw conveyor, it is divided into two scales: a coarse scale into five segments (A, B, C, D, and E), and a fine scale further subdividing each large segment into two sub-segments, for a total of 10 sub-segments. The switching between multi-scale divisions is determined by the amplitude and density of the load distribution sequence change. When the average difference between adjacent measuring points within a 10-second window is greater than 0.06 or the density of abrupt change points exceeds 1.5 per meter, fine-scale analysis is automatically activated. Candidate anomalous segment identification employs a dual-threshold logic of "gradient abrupt change + continuous evolution": if the gradient change amplitude of a segment exceeds the preset gradient abrupt change threshold ΔGth = 0.045, and the evolution rate Vi > 0.010 / s for five consecutive sampling periods, it is judged as a candidate anomalous segment; if it continues to monotonically increase for another three sampling periods, it is upgraded to an anomalous segment.

[0044] After identifying the anomalous section, the analysis is expanded along the upstream and downstream directions centered on it. The expansion criteria are: if an adjacent section satisfies any two of the following conditions, |Gj|≥0.65|Gi|, or Uj≥0.85Ui, or Ccj≤0.92Cci, it is included in the anomalous influence range. The criteria for the propagation boundary section are: the load gradient change amplitude of the adjacent section is lower than the attenuation threshold of 0.018, and the load evolution rate is lower than the stability threshold of 0.004 / s, while the soil discharge continuity recovers to above 0.93. The main expansion direction of the anomalous section is determined based on the direction of the maximum spatial gradient Vi of each section within the influence range; the expansion speed ve is calculated by dividing the offset of the adjacent periodic anomalous boundary position by the time interval, using the unit m / min. The development trend extrapolation prediction uses a combination of exponential smoothing and linear extrapolation, with a smoothing coefficient λ=0.35 and an extrapolation time domain of 20s.

[0045] The dynamic control window generation module calculates the pressure maintenance requirement parameters of the earthen silo based on the equivalent conveying capacity parameter Qeq and the earthen silo pressure change rate dP / dt, and obtains the conveying capacity-pressure demand deviation parameter Dm. Then, based on the magnitude, direction of change, and rate of change of Dm, the current operating state is divided into three categories: stable stage, transitional adjustment stage, and abnormal control stage. Specifically: Stable stage: Dm ∈ [-0.08, 0.12] and |dDm / dt| < 0.02 / min; Transitional adjustment stage: Dm ∈ [-0.20, -0.08) or (0.12, 0.25], or |dDm / dt| between 0.02 and 0.06 / min; Abnormal control stage: Dm < -0.20 or |dDm / dt| > 0.06 / min, or simultaneous abnormal segment expansion. Under different operating stages, the stage control model is invoked respectively. Under the stable stage, the target interval for soil discharge capacity Qtar is Qdem × [0.98, 1.05]. Under the transitional adjustment stage, the continuous mapping function Qtar = Qdem is used. ·[1+0.6·tanh(1.8Dm)]; achieve smooth adjustment; during the abnormal control stage, the amplitude constraint function is adopted: Qtar=min(max(Qdem·[1+1.2Dm],Qmin),Qmax); where Qmin=78m3 / h, Qmax=152m3 / h. Then, based on Qtar and the change rate of earth pressure, the allowable fluctuation range of earth pressure is constructed. In Example 1, the allowable fluctuation range of the stable stage is Pset±0.010MPa, the transition adjustment stage is Pset±0.008MPa, and the abnormal control stage is tightened to Pset±0.006MPa. The allowable adjustment range of tunneling propulsion parameters is determined by the multivariate constraint mapping function: [vj,min,vj,max]=F(Qtar,dP / dt,Pallow,Risk);

[0046] If Qtar < 100 m³ / h, the upper limit of the advance speed will automatically tighten to 32 mm / min; if Qtar ≥ 130 m³ / h and the soil chamber pressure is stable, the advance speed can be widened to 42 mm / min. The advance pressure adjustment range is usually 11.5–14.8 MPa, shrinking to 11.8–13.6 MPa during abnormal stages. The adjustment range for slag soil improvement includes foam expansion ratio (FER), foam injection ratio (FIR), bentonite injection volume (BIR), and polymer injection volume (PIR). During the stable stage, FER is 10–15, FIR is 28–42 L / m³, and BIR is 8–15 L / m³; during the abnormal stage, FER is increased to 14–20, FIR is increased to 45–65 L / m³, and PIR is increased to 1.5–3.0 L / m³ if necessary.

[0047] When the main expansion direction of the abnormal section is clear, the system makes directional corrections through a spatial constraint correction function. When the abnormality expands downstream, i.e., towards the soil outlet, the system prioritizes limiting the sudden increase in soil discharge capacity to prevent "idling and soil dumping—pressure drop." At this time, the upper limit of Qtar is reduced by 8%–12%, the upper limit of propulsion speed is reduced by 10%–15%, while the adjustment range of foam and polymer is increased by 12%–25%. When the abnormality approaches the propagation boundary and shows a decay trend, the system uses a gradual recovery function to adjust the range by 3%–5% per cycle, rather than restoring to the normal window in one step, to avoid secondary disturbances. Furthermore, the system adjusts the range based on the soil discharge intermittent frequency parameters, transport stability parameters, and the output of the risk identification module. Risk indicators are used to construct risk constraint functions, and safety boundaries are corrected for each parameter window. Then, using a historical working condition matching model, working conditions similar to the current soil layer, advance speed, soil chamber pressure, and amount of amendment used are retrieved from the construction database of at least 200 past cycles. The window boundaries are corrected a second time. Finally, the system maps the target range of soil discharge capacity, the allowable fluctuation range of soil chamber pressure, the allowable adjustment range of tunneling advance parameters, and the adjustment range of slag amendment into a multi-parameter collaborative control domain. Through a multi-objective optimization scheduling model, the system performs time-series scheduling and dynamic updates according to the priority of "soil chamber pressure stability first, risk suppression second, transportation efficiency third, and energy consumption optimization last", forming a multi-parameter dynamic collaborative control window.

[0048] refer to Figures 3-5 Example 2: Collaborative regulation of high disturbance risk in water-rich silty fine sand-silt interbedded with local sand and gravel strata

[0049] Project Conditions and Implementation Objectives: This invention is applicable to water-rich silty fine sand and silt interspersed with localized lenses of sand and gravel. In shield tunneling, this type of stratum exhibits large fluctuations in soil fluidity, rapid changes in the compressibility of excavated soil, and the continuity of excavated soil is easily affected by groundwater and coarse particles. The soil chamber pressure is also more sensitive to changes in propulsion and excavation. Compared to Example 1, this example emphasizes closed-loop reconfiguration control, predictive compensation correction, risk identification, and graded control functions under complex working conditions. It is used to illustrate the implementation of this invention in highly volatile and high-risk strata. The project object is an earth pressure balance shield tunneling machine with an excavation diameter of 6.98m, an effective length of 10.8m for the screw conveyor, a blade outer diameter of 900mm, a rated speed of 24rpm, and a drive motor rated power of 200kW. The propulsion system is equipped with 14 sets of propulsion cylinders, with a total rated thrust of 42000kN. The average permeability coefficient of the strata is 1.2×10⁻⁵~4.6×10⁻⁴ m / s, the groundwater level is about 1.5m below the surface, and the overburden thickness is 8.4m~15.2m. The target soil chamber pressure range is 0.20~0.29MPa, and the conventional advance speed is 18~38mm / min. To improve the safety margin under complex working conditions, the system, based on Example 1, adds a microwave moisture content detector at the soil outlet, using an E+HSolitrend MMP44; a Kistler 8763B dynamic pressure sensor is added to the key sections of the screw conveyor casing; the slag improvement system adds a dual-channel foam generator and an online flow closed-loop module, with the foam generator selected being the Putzmeister EP series, with a maximum single-channel liquid supply capacity of 90L / min; the edge processing unit adopts a dual-machine hot standby architecture and is equipped with a UPS uninterruptible power supply to ensure continuous control capability under communication jitter and voltage fluctuation conditions.

[0050] The overall structure of the closed-loop reconfiguration control module: The closed-loop reconfiguration control module is deployed inside the central control unit and consists of a candidate adjustment quantity generation unit, an execution priority calibration unit, a conflict judgment and consistency correction unit, a phased execution unit, a timing decoupling control unit, a hysteresis compensation unit, a prediction compensation unit, and a closed-loop update unit. This module receives the control interval output by the dynamic control window generation module and combines it with real-time deviation parameters, the main expansion direction of the abnormal section, the expansion speed, the risk status, and the current feedback status of the execution object to form specific control instructions for each adjustment cycle. Its goal is not to push all control variables to the boundary values ​​at once, but to progressively adjust them according to the deviation change trend within each 2-second adjustment cycle, so that the soil pressure, soil discharge capacity, and propulsion parameters gradually converge.

[0051] The method for generating the candidate adjustment set: At the beginning of each control cycle, the system generates a candidate adjustment set based on the allowable range of each parameter in the multi-parameter dynamic collaborative control window, namely, screw conveyor speed, propulsion speed, propulsion pressure, foam injection ratio, bentonite injection amount, and polymer injection amount. In Example 2, the candidate set generation rules for each execution object are as follows: For screw conveyor speed ns, if the current allowable range is [nmin, nmax], then the candidate adjustment amount is: Δns∈{-1.2,-0.8,-0.4,0,+0.4,+0.8,+1.2}rpm; and options that cause the target speed to exceed the allowable range are eliminated; For propulsion speed vj, the candidate adjustment amount is: Δvj∈{-6,-4,-2,0,+2,+4}mm / min; For propulsion pressure Pj, the candidate adjustment amount is: ΔPj∈{-0.6,-0.4,-0.2,0,+0.2,+0.4}MPa; For foam injection ratio F For IR, the candidate adjustment amount is: ΔFIR∈{-6,-3,0,+3,+6,+9}L / m^3; for bentonite injection amount BIR, the candidate adjustment amount is: ΔBIR∈{-3,0,+2,+4,+6}L / m^3; for polymer injection amount PIR, the candidate adjustment amount is: ΔPIR∈{0,+0.5,+1.0,+1.5,+2.0}L / m^3. After the candidate adjustment amount set is generated, the system further performs priority calibration based on the transport capacity-pressure demand deviation parameter Dm and the main expansion direction of the abnormal section. If Dm is negative and the abnormality expands in the downstream direction, it indicates that the transport capacity is insufficient and the abnormality is spreading towards the soil discharge end. At this time, the priority order is usually: increase the improvement parameter > decrease the propulsion speed > decrease the propulsion pressure > fine-tune the screw speed. If Dm is positive but the soil chamber pressure is decreasing rapidly, it indicates that there is a risk of over-discharge. The priority is adjusted to: decrease the screw speed > decrease the soil discharge capacity > maintain or slightly decrease the propulsion speed > maintain the improvement parameter.

[0052] To facilitate field application, this embodiment sets the following priority scoring function: Scorek = λ1·Sdev + λ2·Sdir + λ3·Srisk + λ4·Sact; where Sdev represents the score for the ability to correct deviation parameters, Sdir represents the score for matching with the direction of anomaly propagation, Srisk represents the score for safety adaptation under the current risk state, and Sact represents the score for the response speed of the execution object; in Embodiment 2, λ1 = 0.42, λ2 = 0.24, λ3 = 0.22, and λ4 = 0.12. The higher the score, the higher the priority for entering the execution sequence. In the response speed score, foam injection and changes in screw speed usually respond quickly, followed by propulsion pressure, and propulsion speed is the slowest. Therefore, in high-risk phases, the system will prioritize calling fast-response execution objects.

[0053] By first generating multiple candidate adjustment variables and then prioritizing them according to deviation, direction, risk, and response speed, this embodiment avoids the single-action approach of "simply reducing speed or adding bubbles upon detecting an anomaly" in traditional control. Instead, it forms a multi-variable, selectable, comparable, and optimizable adjustment decision-making basis. This improves the adaptability of the control and provides sufficient conditions for subsequent conflict resolution and phased execution.

[0054] Conflict determination and formation of consistent adjustment quantities: Under complex operating conditions, different controlled objects often have opposite effects on the same constraint variable. For example, increasing the screw speed may improve the soil discharge capacity, but it may also cause the soil pressure in the soil chamber to drop; reducing the propulsion speed is beneficial to mitigate pressure fluctuations, but if the screw speed is reduced at the same time, it may cause the soil in the soil chamber to stagnate and increase pressure. Therefore, this embodiment performs conflict determination before execution. When different candidate adjustment quantities have opposite effects on the same target variable, and the magnitude of the opposite effects both exceed the set threshold, the system activates the conflict resolution rules. The specific rules include: First, the safety priority rule: adjustment quantities that are directly related to the current risk level have priority. If the risk of pressure drop in the soil chamber is high, then the adjustment action that would lead to further pressure drop will be suppressed; secondly, the fast response priority rule: when the abnormal expansion rate is greater than 0.35 m / min, the fast response object will be used for initial control first, and then the slow response object will follow up with correction; thirdly, the main effect priority rule: the candidate action that contributes the most to the correction of the deviation parameter will be given priority, and the secondary action will be retained according to the scaling ratio; fourthly, the constraint consistency rule: if the superposition of two adjustment quantities will cause either parameter to go out of the dynamic control window, then the lower priority action will be reduced proportionally or postponed for one cycle.

[0055] The uniform adjustment amount is generated in the following form: Δu'k = ρk·Δuk; where ρk is the uniform correction coefficient, ranging from 0 to 1. If a conflict occurs, the ρk of the low-priority action is reduced in stages by 0.25, 0.5, or 0.75; in the case of a strong conflict, it is directly set to 0. For example, if the candidate actions in a certain cycle are "screw speed +0.8 rpm", "propulsion speed +4 mm / min", and "foam injection +6 L / m^3", and the current state is under pressure drop warning, then the propulsion speed increase action is directly canceled because it conflicts with the pressure maintenance target. The ρk of the screw speed increase action is corrected to 0.5, and only +0.4 rpm is executed. At the same time, the foam injection is executed in full to ensure that the conveying capacity is improved without causing a significant pressure drop. Conflict judgment and uniform correction prevent multiple controlled objects from acting independently and instead form coordinated actions under a unified goal. This mechanism can effectively avoid the mutual cancellation phenomenon of "one module requiring accelerated soil discharge and another module requiring pressure stabilization and emission reduction", thereby reducing control oscillations and improving the stability of joint regulation of multiple execution objects.

[0056] Phased Adjustment and Timing Decoupling Control: To prevent overshoot caused by synchronous operation of all executing objects, this embodiment employs phased adjustment and timing decoupling control. Each standard adjustment cycle is 2 seconds, and a complete phased execution round consists of 3 control sub-stages: First sub-stage (0-2 seconds): Prioritize the execution of parameters related to the screw conveyor capacity, namely screw speed ns, fine-tuning of the discharge gate (if applicable), and foam injection volume FIR. During this stage, the propulsion speed remains constant or only allows gradual changes, and the propulsion pressure is only allowed to be finely adjusted within ±0.1 MPa; Second sub-stage (2-4 seconds): Based on the rate of change of soil chamber pressure and the rate of load evolution after the execution of the first sub-stage, the propulsion speed vj and propulsion pressure Pj are then adjusted. At this point, a holding constraint or a reverse compensation constraint is applied to the screw speed. For example, if the pressure in the soil chamber rises too quickly due to the decrease in propulsion, a small compensation of +0.2 to 0.4 rpm can be made to the screw speed. The third sub-stage (4 to 6 s): Based on the results of the first two stages, the bentonite and polymer parameters are tuned to make the rheological properties of the slag soil closer to the target state and to provide more favorable conveying conditions for the next cycle.

[0057] When the abnormal section exhibits a bridging-type blockage trend in the middle and later stages, the system typically employs the following adjustment sequence: first, increase the injection volume of foam and polymer to improve the plasticity of the excavated soil; then, slightly reduce the propulsion speed; subsequently, depending on the recovery of the soil chamber pressure, decide whether to maintain the screw speed or slightly increase it. If the soil chamber pressure decreases while soil discharge remains intermittent, prioritize reducing the screw speed and slowing down the propulsion, rather than immediately continuing to add foam, to avoid excessive liquid phase leading to an increased risk of gushing. Through this time-sequential decoupling control, "soil discharge capacity adjustment" and "propulsion adjustment" can be separated in time, reducing system coupling excitation caused by simultaneous changes in multiple variables.

[0058] Staged adjustment and time-sequence decoupling control can effectively reduce the coupling impact when multiple variables act simultaneously, making the control process smoother and more predictable. This mechanism is particularly suitable for the characteristics of shield tunneling, such as "fast response of the screw conveyor, slow response of the propulsion system, and delayed effect of the amendment," and can significantly reduce soil pressure overshoot, soil discharge pulsation, and frequent reverse actions of the actuator.

[0059] Rate of Change Limits and Hysteresis Compensation: Due to the significant inertia and execution lag in the shield tunneling system, rapid changes in control parameters can easily lead to drastic fluctuations in soil chamber pressure and soil discharge status. Therefore, this embodiment imposes rate of change limits on each execution object. Specific limits are: screw speed change rate not exceeding 0.8 rpm / 2s, propulsion speed change rate not exceeding 4 mm / min / 2s, propulsion pressure change rate not exceeding 0.35 MPa / 2s, foam injection ratio change rate not exceeding 8 L / m³ / 2s, bentonite injection volume change rate not exceeding 5 L / m³ / 2s, and polymer injection volume change rate not exceeding 1.2 L / m³ / 2s. If the uniform adjustment exceeds the above limits, the system automatically splits the process and executes it incrementally in subsequent cycles.

[0060] Hysteresis compensation is used to offset the time delay between the issuance of the command and the response of the working condition. In Example 2, the average response delay of each executing object is calibrated as follows: screw speed response delay 0.7-1.1s, propulsion pressure response delay 1.2-1.8s, propulsion speed response delay 2.5-4.0s, foam injection state activation delay 2.0-3.5s, and the significant influence of bentonite / polymer on the rheology of the slag soil delay approximately 4.0-8.0s. Therefore, the system establishes a hysteresis compensation term for each object: ucomp(t)=ucmd(t)+Kd·[ypred(t+τ)-yest(t)]; where τ is the object calibration delay and Kd is the compensation coefficient. Kd for screw speed is taken as 0.18, propulsion speed as 0.11, and foam injection as 0.21. By predicting short-term pressure changes and load evolution trends, some anticipated deviations can be pre-incorporated into the current control parameters. This reduces the sluggish response caused by "control measures being issued before the operating conditions have changed." Rate limiting ensures that control parameters do not change too rapidly, while hysteresis compensation reduces the negative impact of execution lag on control effectiveness. The combination of these two methods prevents the system from abruptly changing or slowly spiraling out of control, thereby improving the accuracy of soil chamber pressure control and anomaly suppression efficiency. This approach is particularly suitable for tunnel boring machine (TBM) systems with significant inertia and hysteresis.

[0061] Safety Degradation Triggering and Predictive Compensation Correction: When the abnormal section expansion speed is detected to exceed a preset threshold or the risk indicator reaches a preset threshold, the system triggers safety degradation control. In Example 2, the abnormal expansion speed threshold is set to 0.55 m / min; any one of the three risk indicators—gushing, pressure drop, and blockage—exceeding 0.78, or the coupled risk parameter exceeds 0.72, triggers safety degradation. After safety degradation, the dynamic control window shrinks overall: the upper limit of the screw speed is reduced by 10%–18%, the upper limit of the propulsion speed is reduced by 15%–25%, the propulsion pressure adjustment range shrinks by 20%, and the foam and bentonite parameter windows shift towards the higher modification side. At the same time, the system applies amplitude constraints to all execution objects, prohibiting them from exceeding the upper boundary of the degradation window.

[0062] Within each adjustment cycle, the predictive compensation unit makes feedforward corrections to the target adjustment amount for the next cycle based on the changing trends of the abnormal section expansion rate ve and the deviation parameter Dm. Example 2 uses the following predictive compensation formula: Δu{k+1}^{pre}=Δuk+a1·ΔDm+a2·Δve+a3·Risk' where ΔDm is the deviation change in this cycle, Δve is the change in the abnormal expansion rate, Risk' is the risk trend derivative term, and a1, a2, and a3 are taken as 0.45, 0.33, and 0.22, respectively. If the risk continues to rise, conservative control components are added in advance for the next cycle, such as prioritizing tightening the upper limits of advance and soil removal; if the risk decreases and the abnormal boundary begins to shrink, the compensation intensity is gradually reduced. Through safety degradation and predictive compensation, the system no longer only reacts to "already occurred deviations," but can take measures in advance based on the rising risk and abnormal expansion trends. This forward-looking closed-loop control method can effectively compress the risk development time window and reduce the probability of severe operating conditions.

[0063] Operating parameter re-acquisition, convergence judgment, and closed-loop update: After each adjustment cycle, the system re-acquires all operating parameters, recalculates the equivalent conveying capacity parameter Qeq and the conveying capacity-pressure demand deviation parameter Dm, and compares them with the preset convergence conditions. In Example 2, the convergence conditions are set as follows: for three consecutive adjustment cycles, the following conditions must be met: |Dm|≤0.08, |dP / dt|≤0.003MPa / s, abnormal expansion rate ve≤0.08m / min, and soil discharge continuity coefficient Cc≥0.92. If the convergence conditions are not met, the step size is adaptively adjusted according to the deviation change amplitude and change rate. The step size adjustment rule is: γ{k+1}=γk·[1+b1·|Dm|+b2·|dDm / dt|]; where b1=0.35, b2=0.22; but the upper limit of γ does not exceed 1.4, and the lower limit is not lower than 0.6. If the deviation is decreasing, the step size is gradually reduced to prevent overshoot when approaching steady state; if the deviation is increasing, the step size is appropriately increased to enhance the correction force. After the convergence condition is met, the system maintains the current operating parameters and enters steady-state maintenance mode, making only minor adjustments to small disturbances and no longer performing large-step incremental updates.

[0064] Risk Protection and Control Module: Establishment of Risk Indicator System: The risk protection and control module is specifically designed to identify surge risk, soil chamber pressure drop risk, and transport blockage risk, and constructs multi-risk coupling parameters. The three types of risk indicators are defined as follows:

[0065] The surge risk indicator Rsurge mainly consists of the absolute value of soil chamber pressure, pressure fluctuation amplitude, soil moisture content, excessive foam injection status, and abnormal instantaneous flow rate at the soil outlet. Example 2 uses: Rsurge = 0.28·P* + 0.22·Ap + 0.18·Wo^ + 0.17·FIR* + 0.15·Qout*. The soil pressure drop risk index Rdrop is mainly composed of the target pressure difference, pressure drop rate, advance speed, degree of over-discharge capacity, and soil pressure imbalance coefficient: Rdrop = 0.30·ΔP* + 0.25·|dP / dt| + 0.18·vj^ + 0.15·Qexcess* + 0.12·Up*. The transport blockage risk index Rblock is mainly composed of the section load gradient, load evolution rate, continuous decrease in discharge rate, unit volume energy consumption, and coarse particle identification factor: Rblock = 0.26·|G|* + 0.24·|V| + 0.18·(1-Cc)^ + 0.17·Ev* + 0.15·Cg*.

[0066] After normalization, the three risk indicators all range from 0 to 1. Further, based on correlation analysis, a multi-risk coupling parameter Rcpl is constructed:

[0067] Rcpl = μ1Rsurge + μ2Rdrop + μ3Rblock + μ4Cov(Rsurge, Rdrop, Rblock), where μ1, μ2, μ3, and μ4 are 0.24, 0.30, 0.30, and 0.16, respectively. The Cov term is used to characterize the degree of coupling reinforcement when multiple risks rise together. Dynamic risk threshold update and risk level classification: Since the risk boundaries are different in different strata and different improvement states, this embodiment does not use a fixed risk threshold, but rather updates it adaptively based on the statistical distribution characteristics of operating parameters and historical tunneling data. A dual-time-scale sample set is constructed using the operating data of the most recent 50 rings and the most recent 20 minutes. The mean, standard deviation, and 95th percentile of each risk indicator are calculated, and then the risk threshold is updated in combination with the historical successful working condition database. In Embodiment 2, the dynamic risk threshold update formula is: Thi = θ1·Q95i + θ2·Meani + θ3·Histi; where θ1 = 0.5, θ2 = 0.2, and θ3 = 0.3. The multi-risk coupling parameter Rcpl is then compared with the dynamic risk threshold, and a risk development trend parameter Tr is used for classification. The risk development trend parameter is defined as the extrapolated risk growth rate over the next 10 seconds. The system classifies risk states into four levels: Normal state: Rcpl < 0.35 and Tr ≤ 0; Warning state: 0.35 ≤ Rcpl < 0.55 or Tr > 0 but is relatively small; Risk state: 0.55 ≤ Rcpl < 0.75, and Tr is continuously increasing positively; Dangerous state: Rcpl ≥ 0.75 or any single risk ≥ 0.85. The dynamic risk threshold avoids the problem of fixed thresholds being either too conservative or slow to react in complex geological formations, enabling risk assessment to be adjusted in real time according to on-site conditions, thereby improving the accuracy and adaptability of risk identification.

[0068] Window contraction control under warning and risk states: When the system is in a warning or risk state, the risk protection control module embeds multiple risk coupling parameters as constraint variables into a multi-parameter dynamic collaborative control window. Example 2 uses the risk constraint mapping function: W'j = Wj·(1-κ·Rcpl·φ(Tr)); where Wj represents the original control window width, W'j is the corrected width, κ is the contraction coefficient (0.18–0.36 in Example 2), and φ(Tr) is the trend gain function (1.2–1.5 if the risk continues to rise, and 1.0 if it tends to stabilize). Through this function, the system can simultaneously shrink the target range of soil discharge capacity, the allowable fluctuation range of soil chamber pressure, and the allowable adjustment range of propulsion parameters. When Rcpl=0.61 and Tr continues to increase positively, the allowable range of screw speed can shrink from [8.0,11.5] rpm to [8.2,10.1] rpm, the propulsion speed from [24,36] mm / min to [22,29] mm / min, and the allowable fluctuation range of soil chamber pressure from ±0.008MPa to ±0.005MPa. At the same time, the foam injection and bentonite parameter windows shift towards higher modification to enhance the plasticity and sealing of the excavated soil. The window shrinkage control allows the system to enter a more conservative and robust working mode in advance when the risk increases but has not yet reached a dangerous state, thereby suppressing the spread of risk without complete shutdown and improving construction continuity and safety.

[0069] Graded safety degradation control under hazardous conditions: When entering a hazardous state, this embodiment triggers graded safety degradation control. The degradation control is executed according to a preset degradation sequence, in three levels: Level 1 degradation: The upper limit of soil discharge capacity is immediately reduced by 12% to 18%, the upper limit of propulsion speed is reduced by 15%, and the foam injection ratio is increased by 15% to 25%; Level 2 degradation: If the risk does not decrease within two cycles, the propulsion pressure is further limited to near the lower safety limit, the amount of bentonite and polymer injected is increased by 20% and 30% respectively, and the screw speed is subject to a limit constraint to avoid drastic speed increases; Level 3 degradation: If the risk continues to expand, a safe operation mode is entered, the propulsion speed is reduced to the minimum safe value, the screw speed is maintained in the low-speed range with priority given to pressure maintenance, the improved parameters are maintained at a high level, and any acceleration actions that may amplify fluctuations are prohibited. During the degradation process, each execution object is simultaneously subject to linkage coordination constraints and change rate limits. For example, it is not allowed to rapidly increase the screw speed while the propulsion speed drops significantly, so as to avoid short-term depletion of the soil chamber pressure; nor is it allowed to continue to significantly increase bentonite and polymer in the same cycle when foam increases significantly, so as to prevent excessive fluidization of the slag.

[0070] The tiered safety downgrade control system transforms the handling of hazardous conditions from "single-time forced suppression" to "layered progressive contraction," ensuring safety while preserving as much construction continuity as possible. Its advantage lies in the ability to escalate measures step-by-step based on whether the risk continues to increase, avoiding unnecessary over-conservative control.

[0071] Dynamic Correction and Further Degradation Control: During the implementation of safety degradation, the system does not rigidly execute preset actions, but rather corrects the degradation strategy in real time based on multiple risk coupling parameters and risk development trend parameters. If the risk of gushing decreases while the risk of blockage increases, the system focuses on reducing the helical load and improving performance without continuing compression propulsion. If the risk of pressure drop increases the fastest, pressure preservation is prioritized, further compressing the soil discharge capacity and propulsion speed. If a continuous increase in risk is detected, further degradation control is triggered, i.e., switching from the current degradation level to a more stringent level. In Example 2, the further degradation judgment condition is: the increase in Rcpl is greater than 0.06 in two consecutive adjustment cycles, or any single risk exceeds 0.90. At this time, the system can directly skip the intermediate recovery step and enter a higher-level safe operation mode.

[0072] Phased rollback and gradual recovery after risk subsidence: When the risk parameters coupled with multiple risks decrease below the recovery threshold, the system executes phased rollback and recovery scheduling control. In Example 2, the recovery threshold is set as follows: Rcpl ≤ 0.42 and no longer increases for three consecutive cycles, while all individual risks are less than 0.50. The recovery process also adopts a phased approach: First recovery phase: Firstly, the limitation on soil discharge capacity is removed, but the partial tightening of the advance speed and advance pressure is retained; Second recovery phase: If stable for three consecutive cycles, the advance parameters and improvement parameter windows are gradually widened, with each cycle widening by 15% to 20% of the original shrinkage; Third recovery phase: When the deviation parameters, soil pressure change rate, soil discharge continuity, and abnormal expansion speed all return to normal, the control window gradually returns to the conventional collaborative control mode.

[0073] Consistency constraints are applied during the recovery process, prohibiting the recovery of a single parameter to the baseline condition in one step. For example, when the screw speed is recovered, the propulsion speed must be recovered after a sub-stage. The recovery rate of foam and polymer must not be faster than the propulsion recovery rate to prevent the slag that has just been stabilized from losing its flow equilibrium again.

[0074] In the aforementioned hazardous situation case, after Rcpl dropped from 0.79 to 0.39 and remained there for 12 seconds, the system began to recover. First, the upper limit of the screw speed window was restored to 0.5 rpm, but the propulsion speed was still controlled below 26 mm / min. Then, in two cycles, the upper limit of the propulsion speed was gradually restored to 30 mm / min, while the foam injection ratio was slowly reduced from 63 L / m³ to 54 L / m³, and then the bentonite and polymer were gradually reduced. The entire recovery process lasted approximately 22 seconds, during which the pressure in the soil chamber remained stable without any further abrupt changes.

[0075] refer to Figures 5-8 Example 3: Applicable to engineering projects involving long distances, stiff plastic clay interbedded with sand layers, and requiring high reliability and fault tolerance control.

[0076] Project Background and Implementation Objectives: This project is suitable for long-distance shield tunneling, especially for projects with significant longitudinal geological variations, thin layers of stiff plastic clay interbedded with silt in some sections, localized medium-coarse sand lenses, and high requirements for automated and reliable continuous operation in construction organization. The shield machine used in this embodiment is an earth pressure balance shield machine with an excavation diameter of 7.52m, an effective screw conveyor length of 11.4m, a rated maximum speed of 26rpm, a rated power of 250kW for the drive motor, a total thrust of 46000kN for the propulsion system, and a rated torque of 5600kN·m for the cutterhead. The section is 2.86km long, with an average burial depth of 17.8m and a maximum burial depth of 24.3m. The geological strata are mainly stiff plastic clay, with localized medium-dense sand layers and a small amount of gravel interlayers. Groundwater pressure varies significantly along the tunnel. Compared with the previous two embodiments, this embodiment places greater emphasis on system-level architecture reliability, control link redundancy, inter-module time synchronization, fault-tolerant degradation takeover, and a complete closed-loop execution process at the method level, so as to illustrate that the present invention can not only achieve collaborative control at the algorithm level, but also be stably deployed in the engineering site in the form of an industrial control system.

[0077] The implementation objectives include: First, establishing a two-layer control architecture composed of a central control unit, an edge processing unit, and a data communication module to ensure data consistency and execution stability during multi-module collaborative operations; Second, constructing a collaborative constraint matrix among the screw conveyor drive device, propulsion device, and slag improvement device to enable the executed objects to implement linkage adjustment under unified commands; Third, providing an alternative implementation method for the equivalent conveying capacity parameters, namely, introducing the standard deviation of section load, time sequence memory term, state discrimination gating processing, and adaptive updating of nonlinear mapping function; Fourth, fully outlining the implementation steps of the collaborative control method.

[0078] Central Control Unit, Edge Processing Unit, and Data Communication Module: Control Architecture: The system adopts a three-tiered collaborative architecture of "Central Control Unit + Edge Processing Unit + Data Communication Module". The Central Control Unit is located in the supporting control room behind the tunnel boring machine, using two industrial servers to form a primary-backup redundant architecture. The server model is Advantech MIC-770V3, with an Intel Xeon W-1290E processor, 64GB of memory, 2TB of solid-state storage, and the operating system is Ubuntu LTS real-time kernel version. The control platform is deployed using containerized services, and the various algorithm modules communicate through an internal message bus. The Central Control Unit communicates with the multi-segment data acquisition module, the process feature construction module, the equivalent conveying capacity calculation module, the conveying status and capacity determination module, the dynamic control window generation module, the closed-loop reconfiguration and control module, and the risk protection control module, respectively, to execute global scheduling control, multi-parameter dynamic collaborative control window generation, and final control command arbitration.

[0079] The edge processing unit is located between the multi-segment data acquisition module and the central control unit. Two units are installed along the main section and the auxiliary section of the tunnel boring machine (TBM). The edge processing unit is an Advantech UNO-238V2, equipped with an Intel Atomx 6425E processor, 16GB of RAM, and a 512GB industrial-grade SSD. The edge processing unit primarily handles tasks such as real-time data filtering, outlier removal, preliminary feature extraction, communication caching, and local rapid response control. For data streams with a sampling period of 200ms, the edge unit first performs 5-point median filtering and first-order low-pass filtering before adding a unified timestamp to the results and sending them to the central control unit. When the edge processing unit detects that the soil pressure change rate exceeds 0.010MPa / s for three consecutive sampling periods, or that the absolute value of the load gradient in a critical section of the screw conveyor exceeds 0.055 for five consecutive sampling periods, it can trigger local rapid response control, such as limiting the screw speed change or freezing the propulsion speed increase command, to shorten the response time to sudden anomalies.

[0080] The data communication module includes independent data transmission channels and control command channels. The data transmission channel adopts a hybrid structure of PROFINET IRT gigabit industrial Ethernet backbone and EtherCAT branch, with dual-port ring network redundancy for key measurement points. The control command channel uses an independent industrial control bus channel, sending commands to devices including screw conveyor frequency converters, propulsion cylinder proportional valve groups, foam pump stations, bentonite grouting pumps, and polymer metering pumps. To prevent data congestion from affecting real-time control, this embodiment completely decouples the uploading of operating parameters and the sending of control commands. The data channel refresh cycle is 200ms, the control command channel refresh cycle is 500ms, and the risk alarm channel has an independent refresh cycle of 100ms.

[0081] Dual-plane architecture and time synchronization: The central control unit employs a dual-plane architecture consisting of a "data processing plane" and a "control execution plane." The data processing plane is primarily responsible for feature construction along the route, equivalent transport capacity calculation, abnormal section identification, risk indicator calculation, and historical operating condition matching. The control execution plane is primarily responsible for dynamic control window generation, closed-loop reconstruction and control, safety degradation control, and control command arbitration. The two planes interact through a shared memory queue and event triggering interface, but their respective threads run independently to avoid high-frequency data calculations blocking control issuance. The main refresh cycle of the data processing plane is 1 second, the main refresh cycle of the control execution plane is 2 seconds, and the risk emergency sub-thread operates independently at 0.5 seconds.

[0082] To ensure consistency in multi-module collaborative computing, the central control unit performs unified time synchronization processing on data from different modules and establishes a causal correspondence between operating parameters and control commands based on timestamps. This embodiment configures a Meinberg LANTIME M300 time server to provide unified time synchronization for each edge unit, PLC, industrial camera, and acquisition card via the PTP protocol, keeping the system's internal clock deviation within ±5ms. All operating parameters, calculation results, and control commands are timestamped at the millisecond level, and a four-element association index of "acquisition time—processing time—execution time—response confirmation time" is established in the database. For example, if an abnormal spiral load at a certain time t0 causes a control window to be generated at t0+2s, the system can accurately trace the data segment and risk status on which the window is based, avoiding misjudgments caused by misaligned data timing.

[0083] Control command consistency arbitration and multi-level buffer: Under complex operating conditions, the dynamic control window generation module, closed-loop reconfiguration control module, and risk protection control module may simultaneously output control suggestions. To prevent command conflicts, the central control unit sets up a consistency arbitration mechanism. The arbitration priority rule is as follows: in dangerous conditions, the risk protection control module has the highest priority, followed by the closed-loop reconfiguration control module, and finally the dynamic control window regular optimization module; in non-dangerous conditions, the closed-loop reconfiguration control module takes precedence over the window optimization module. If the spiral speed adjustment directions output by the two modules are opposite, the instruction from the higher-priority module takes precedence, and the lower-priority instruction is reduced or postponed proportionally; if the instructions have the same direction of action on the same object, weighted fusion is performed. The weighted fusion formula is: uout=ξ1urisk+ξ2uloop+ξ3uwindow; where ξ1, ξ2, and ξ3 are dynamically set according to the current risk level and deviation degree, and can be 0.6, 0.3, and 0.1 in dangerous conditions, and 0.2, 0.3, and 0.5 in steady-state conditions.

[0084] The central control unit also establishes multi-level data caches among the modules. The first-level cache stores high-frequency raw operating parameters from the last 5 minutes for anomaly tracing; the second-level cache stores feature calculation results and risk indicators from the last 3 hours for short-term historical condition matching; and the third-level cache stores historical construction data for the entire line, including tunneling parameters for each ring, improvement parameters, alarm records, and manual handling records, for adaptive updates of nonlinear mapping functions and dynamic correction of risk thresholds. The caches employ a hybrid architecture of TimescaleDB time-series database and local memory. The first-level cache resides in memory in a circular queue, while the second and third-level caches are written to SSDs.

[0085] The consistency arbitration mechanism ensures that multiple modules do not conflict when participating in control at the same time. The multi-level buffer allows the system to utilize high-frequency current data and absorb historical experience, thereby achieving coordinated control that balances "real-time response" and "historical self-adaptation". Fault-tolerant takeover under communication and central control anomalies: When the data communication module malfunctions or the central control unit malfunctions, the edge processing unit implements fault-tolerant control according to the preset degradation takeover strategy chain. The takeover logic is divided into three levels: Level 1 Local Control: When the communication interruption of the central control unit lasts for more than 1 second but less than 5 seconds, the edge unit maintains the most recent stable control window and only allows the screw speed and propulsion speed to change slowly within a small range, prohibiting large adjustments; Level 2 Limiting Control: When the interruption lasts for more than 5 seconds or the central control unit loses its heartbeat for more than 3 cycles, the edge unit restricts all execution objects to a conservative safety window, for example, the propulsion speed is locked within 85% of the current value, the screw speed change is limited to ±0.4 rpm, and the foam injection ratio is maintained at more than 10% higher than the normal median; Level 3 Safe Operation Control: When the interruption lasts for more than 15 seconds, or the edge unit detects a rapid abnormal change in pressure, it enters the safe operation mode, the propulsion speed is reduced to the lowest stable value, the screw speed is switched to the pressure-maintaining priority low-speed range, and the modified parameters are switched to the pre-stored safe formula until the central control unit recovers.

[0086] To achieve the above functions, a heartbeat packet is set between the central and edge units every 500ms, and three consecutive lost heartbeat packets are considered abnormal; the control command channel and data channel are each equipped with CRC consistency verification; the edge unit is pre-set with the three most recent stable operating condition templates and two safe operating condition templates as the basis for quick invocation during takeover. The fault-tolerant takeover mechanism enables the present invention to be effective not only under normal network conditions, but also to maintain basic stable operation under common adverse conditions in engineering sites such as communication anomalies and host computer failures, thereby significantly improving system availability and engineering security.

[0087] Execution Object and Control Interface Module: Screw Conveyor Drive Unit: Includes a frequency converter drive unit, a multi-stage speed regulation unit, and a speed change rate constraint unit. The frequency converter drive unit uses an ABB ACS880 series frequency converter, supporting both torque control and speed closed-loop dual modes. The multi-stage speed regulation unit switches between different regulation modes at different operating stages: a fine-grained continuous speed regulation mode with a step size of 0.1 rpm is used in the steady-state stage; an enhanced continuous speed regulation mode with a step size of 0.2 rpm is used in the transition regulation stage; and a limited-amplitude rapid speed regulation mode with a step size of 0.4 rpm is used in the abnormal control and safety degradation stage, but the total change is constrained by the rate. The speed change rate constraint unit is used to limit the screw speed gradient. In this embodiment, the normal maximum change rate is 0.8 rpm / 2s, which decreases to 0.5 rpm / 2s under dangerous conditions. The frequency converter simultaneously reads the motor current, output frequency, and estimated torque, and returns the execution feedback to the central control unit.

[0088] Tunneling propulsion system: The tunneling propulsion system includes propulsion cylinders, a propulsion pressure regulating unit, and a propulsion speed coordination unit. The propulsion cylinders consist of 14 sets of hydraulic cylinders, each with a cylinder diameter of 320mm, a stroke of 2300mm, and a rated working pressure of 31.5MPa. The propulsion pressure regulating unit uses a proportional servo valve group, with the Bosch Rexroth 4WRPEH series as an option. The propulsion speed coordination unit adjusts the speed in tandem with the target propulsion speed issued by the central control unit based on changes in the auger speed. The propulsion speed is not independently controlled but is constrained by soil removal capacity, soil chamber pressure, and risk conditions. For example, when the auger's soil removal capacity is reduced, the propulsion speed coordination unit automatically lowers the upper limit of the target propulsion speed; when the soil chamber pressure is high and the delivery is stable, the propulsion speed is allowed to increase slowly within the control window. The preferred accuracy for propulsion speed adjustment is ±1mm / min, and the preferred accuracy for propulsion pressure adjustment is ±0.05MPa.

[0089] Slag Improvement Unit: The slag improvement unit includes an amendment delivery unit, a flow regulation unit, and an improvement parameter feedback unit. The foam system uses a dual-path foam generator with an adjustable foaming ratio (FER) of 8–22 and an adjustable foam injection ratio (FIR) of 15–80 L / m³. The bentonite system uses a screw pump with an injection rate of 5–30 L / m³. The polymer system uses a precision metering pump with an injection rate of 0–5 L / m³. The flow regulation unit employs closed-loop mass flow control, using E+H Promass series flow meters for foam and KROHNE OPTIFLUX series flow meters for bentonite. The improvement parameter feedback unit collects injection pressure, instantaneous flow rate, cumulative flow rate, valve opening, and mixed liquor concentration in real time and feeds this data back to the central control unit. The control system can adopt different priority adjustment strategies for different modifiers according to the current control window requirements. For example, under the risk of high water content silt, bentonite and polymer are adjusted first, and under the trend of agglomeration and blockage of hard plastic clay, the foam injection ratio is increased first. By treating the screw drive, propulsion system and slag improvement system as controllable execution objects at the same time, this embodiment breaks the separation of "screw control alone, propulsion control alone, and improvement relying on experience" in the traditional system, so that multiple execution objects can be linked and adjusted around the same control window, thereby significantly improving the collaborative control effect.

[0090] Control interface module, collaborative constraint matrix, and multi-source feedback fusion: The central control unit connects to each actuator through the control interface module. The control interface module includes a signal conversion unit, a signal isolation unit, a redundant communication unit, and a dual-channel drive unit. The signal conversion unit converts the host computer's digital control commands into inverter control words, proportional valve analog outputs, and metering pump flow setting signals. The signal isolation unit enhances anti-interference capabilities, preventing high-power equipment of the tunnel boring machine from affecting the control signals. The redundant communication unit provides a main channel and a backup channel; under normal circumstances, the main channel is used, and automatic switching occurs in case of communication failure. The dual-channel drive unit provides main and backup drives for critical actuators to meet the continuity requirements of critical actions.

[0091] The central control unit internally constructs an execution object collaborative constraint matrix Mc to describe the coupling relationship between the screw conveyor speed, propulsion parameters, and improvement parameters. In Embodiment 3, this matrix can be represented as:

[0092] ;

[0093] The columns of the matrix represent, in order, the screw speed, propulsion speed, propulsion pressure, foam injection ratio, and bentonite / polymer combined modification factor. Matrix elements indicate the direction and strength of the coupling effect of one variable on the control effect of another. Before issuing control commands, the central control unit modifies the original control vector based on this constraint matrix to prevent excessive changes in a particular execution object from disrupting the overall balance.

[0094] The control interface module simultaneously receives feedback signals from various actuators and integrates feedback such as speed, current, valve position, flow rate, pressure, and actual propulsion speed through a multi-source feedback fusion unit. A consistency verification unit then determines the reliability of the feedback. If the deviation between the target command and the actual state of an actuator exceeds a threshold—for example, if the difference between the target screw speed and the measured speed is greater than 0.5 rpm and persists for three cycles—the system identifies execution lag or abnormal mechanical load, generating execution state parameters for the upper-level module to correct the control. Through the control interface module and the collaborative constraint matrix, the originally independent device control is transformed into a systemic control constrained by a unified matrix. Multi-source feedback fusion and consistency verification ensure that the control results can be verified and corrected in a timely manner, further improving execution accuracy and control reliability.

[0095] Nonlinear Compensation Correction and Fault Isolation Switching: The central control unit uses a nonlinear compensation mapping function to compensate and correct the control commands based on the deviation between the execution state parameters and the target control commands, in order to reduce the impact of execution lag, nonlinear response, and load changes on the control effect. This embodiment uses the following compensation mapping form: u{cmd}{new}=u{cmd}+c1e+c2e2\operatorname{sgn}(e)+c3\dot{e}; where e is the deviation between the target and the actual execution state, \dot{e} is the deviation change rate, and c1, c2, and c3 vary depending on the execution object. For the screw drive device, they can be 0.42, 0.08, and 0.15; for the propulsion speed coordination unit, they can be 0.30, 0.05, and 0.10; and for the foam flow regulation, they can be 0.36, 0.06, and 0.12. Nonlinear compensation can enhance the correction force when the load changes significantly and reduce the adjustment force when approaching the target, avoiding overshoot.

[0096] When the control interface module detects an anomaly in any actuator or communication channel, the fault isolation unit isolates the abnormal execution path and switches the dual-channel drive unit to the backup drive channel, while simultaneously implementing amplitude limiting control on the corresponding execution object. For example, when the CRC error of the main communication channel of the screw inverter occurs five times consecutively, the system switches to the backup channel within 300ms and freezes the screw speed change within ±0.2rpm of the current value; when the flow feedback of the foam pump station is lost, the system first isolates the automatic closed loop and switches to a semi-automatic mode that operates according to the preset safe flow rate; when the propulsion speed feedback is abnormal, the system maintains the propulsion pressure without increasing and freezes the propulsion speed increase command.

[0097] The method for coordinated control of excavated soil transportation and soil chamber pressure during shield tunneling includes: Step 1: Operational parameter acquisition and preprocessing: During shield tunneling, continuously collect soil chamber pressure, drive load parameters of each axial section of the screw conveyor, energy consumption parameters, and soil discharge continuity parameters. The preferred sampling frequency for soil chamber pressure is 5Hz to 20Hz, the preferred sampling frequency for screw load, current, speed, and vibration parameters is 10Hz, and the preferred sampling frequency for belt scale and soil discharge continuity image recognition data is 5Hz. After acquisition, filtering, outlier removal, and time synchronization are performed. Filtering can use a combination of median filtering and Kalman filtering; outlier removal can use the 3σ rule and abrupt change rule; and time synchronization uses unified timestamp alignment.

[0098] Step 2: Construction of Load Distribution Sequence and Temporal Characteristics along the Helical Conveyor Channel: Based on the driving load parameters, a load distribution sequence along the helical conveyor channel is constructed, and the non-uniformity coefficient of the load in each section, the rate of load change, and the intensity of load fluctuation are calculated. Furthermore, a temporal memory term reflecting the temporal evolution characteristics is constructed using a sliding time window. The time window can preferably be 10s, 15s, or 20s, and the specific value can be set according to the formation change rate and sampling frequency.

[0099] Step 3: Calculation of Energy Consumption and Discharge Continuity Parameters: Calculate the unit volume transport energy consumption parameter based on the energy consumption parameter and the change in discharge volume, and calculate the discharge interruption frequency parameter based on the discharge continuity parameter. The unit volume transport energy consumption can be obtained by dividing the cumulative energy consumption of the current cycle by the discharge volume of the current cycle. The discharge interruption frequency can be determined by combining image recognition and belt scale data.

[0100] Step 4: Calculation of equivalent transport capacity parameters: Based on the section load non-uniformity coefficient, time-series memory term, unit volume transport energy consumption parameter, and soil discharge discontinuity frequency parameter, construct transport resistance characterization parameters and transport capacity characterization parameters, and calculate the equivalent transport capacity parameter Qeq through a capacity-resistance coupling model. The capacity-resistance coupling model can adopt the aforementioned nonlinear mapping model and be updated online based on historical data.

[0101] Step 5: Deviation Parameter Calculation and Phased Division of Operating Status: Based on the relationship between the equivalent conveying capacity parameter and the rate of change of earth chamber pressure, calculate the conveying capacity-pressure demand deviation parameter Dm, and divide the current operating status into phases according to its magnitude, direction of change, and rate of change, including at least a stable phase, a transitional adjustment phase, and an abnormal control phase. If the absolute value of Dm is small and changes slowly, it is a stable phase; if Dm exceeds the normal zone but does not reach a severe mismatch, it is a transitional adjustment phase; if Dm is severely mismatched or accompanied by the expansion of abnormal sections, it is an abnormal control phase.

[0102] Step 6: Generation and Directional Correction of Multi-Parameter Dynamic Cooperative Control Window: Based on the current operating status and the results of abnormal section identification, a multi-parameter dynamic cooperative control window is generated, and the control commands are restricted to the control window range using a constraint projection method. Simultaneously, the control window is directionally corrected according to the direction of abnormal section expansion. For example, when the abnormality expands downstream, the upper limit of the soil discharge capacity is tightened first, and the improvement parameter window is increased; when the abnormality expands upstream, the advance growth is limited first, and the further increase in front-end load is suppressed.

[0103] Step 7: Coordinated Adjustment under Multi-Objective Optimization Scheduling: Under different operating stages, the screw conveyor speed, tunneling propulsion parameters, and slag improvement parameters are coordinated and adjusted based on the multi-objective optimization scheduling model. The multi-objective optimization scheduling model must simultaneously constrain the soil chamber pressure stability, conveying efficiency, and energy consumption level, and can include risk level as an additional constraint. An example of the objective function is as follows: J = \alpha1Jp + \alpha2Jq + \alpha3Je + \alpha4Jr; where Jp represents the pressure stability objective, Jq represents the conveying efficiency objective, Je represents the energy consumption objective, and Jr represents the risk suppression objective. α1 to α4 can be dynamically set according to the operating stage; during abnormal control stages, the weights of Jp and Jr can be increased.

[0104] Step 8: Timing Decoupling and Rate Constraint Execution: During the adjustment process, timing decoupling control is applied to the soil dumping capacity adjustment and tunneling advance adjustment, and change rate constraints and conflict coordination constraints are applied to each control parameter. That is, fast-response objects are executed first, followed by slow-response objects; candidate actions with opposite directions are corrected for consistency; and upper limits are set on the change rate of each object.

[0105] Step 9: Predictive Compensation Correction: In each adjustment cycle, based on the trend of deviation parameter changes and the expansion rate of abnormal sections, predictive compensation correction is performed on the control parameters for the next adjustment cycle, thereby incorporating future trends into the current control decision in advance.

[0106] Step 10: Closed-Loop Update, Adaptive Step Size, and Digital Twin Verification: During the adjustment process, updated operating parameters are collected in real time, and the equivalent transport capacity parameters and deviation parameters are recalculated based on the updated operating parameters to execute closed-loop adjustment. Simultaneously, the adjustment step size is adaptively adjusted according to the magnitude and rate of change of the deviation parameters, and the adjustment results are verified using a digital twin model to meet preset convergence conditions. The digital twin model can construct simplified simulations based on historical loop data and current parameters to predict the evolution trend of soil chamber pressure and load in the next cycle, thereby verifying the rationality of the current control strategy.

[0107] Step 11: Multi-risk coupling classification and safety control: Construct multi-risk coupling risk parameters based on the gushing risk index, the earthwork pressure drop risk index, and the transport blockage risk index, and classify the system operating status based on these risk parameters. Under risky conditions, shrink the multi-parameter dynamic collaborative control window; under dangerous conditions, implement graded safety degradation control.

[0108] Step 12: Dynamic Correction and Gradual Recovery: During the risk control process, the control strategy is dynamically corrected according to the risk development trend. When the risk indicators are reduced to below the recovery threshold, a phased rollback and gradual recovery are performed. At the same time, consistency constraints are applied to the recovery process to ensure that the system smoothly switches to the normal collaborative control mode.

[0109] Working Principle: First, multi-source sensors deployed at the shield tunnel's soil chamber, multiple sections along the axial direction of the screw conveyor, and the soil outlet continuously acquire operating parameters such as soil chamber pressure, pressure change rate, section drive load, conveying energy consumption, and soil discharge continuity. Then, the process feature construction module processes these operating parameters to generate load distribution parameters, load gradient parameters, load evolution rate parameters, and conveying stability parameters that reflect the internal state of the screw conveyor. Next, the equivalent conveying capacity calculation module performs capacity-resistance coupling calculations on the section load characteristics, energy consumption characteristics, and soil discharge continuity characteristics to obtain the overall equivalent conveying capacity parameters. Then, the conveying state and capacity determination module calculates the pressure maintenance requirement parameters based on the soil chamber pressure change rate and determines the degree of mismatch between conveying capacity and pressure requirement. Based on this, the dynamic control window generation module generates parameters according to the current conveying state category, development trend, and... Based on the degree of mismatch, risk status, and historical operating conditions, a multi-parameter dynamic collaborative control window is generated in real time, including the allowable range of soil dumping capacity, the allowable fluctuation range of soil silo pressure, the allowable range of propulsion parameters, and the adjustment range of soil improvement parameters. When the operating parameters deviate from this control window, the closed-loop reconfiguration control module further identifies abnormal sections and their propagation directions, and performs phased, time-decoupled collaborative adjustments to the screw speed, propulsion speed, propulsion pressure, and improvement parameters according to a preset sequence. Based on the updated operating parameters for each cycle, the capacity parameters and deviation parameters are recalculated, and the control actions are progressively corrected. When risks such as gushing, soil silo pressure drop, or conveying blockage are detected, the risk protection control module embeds risk constraints into the control window, suppresses risk expansion through window contraction, linkage limiting, and graded safety degradation control, and gradually restores the system to a normal collaborative control state after the risk is reduced. In summary, this invention uses "multi-segment perception—line modeling—capacity assessment—state determination—control window generation—closed-loop collaborative adjustment—risk protection and recovery" as its core link. Through multi-parameter linkage and continuous feedback updates, it achieves long-term stable collaborative control of excavated soil transportation and soil chamber pressure during shield tunneling construction.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A system for coordinated control of excavated soil transportation and soil chamber pressure during shield tunneling construction, characterized in that, include: Multi-segment data acquisition module: used to collect operating parameters at multiple intervals along the axial direction of the shield tunnel soil chamber, the screw conveyor, and the soil outlet. The operating parameters include: soil chamber pressure and its time change rate, drive load parameters corresponding to each axial position of the screw conveyor, energy consumption parameters during the conveying process, and soil discharge continuity parameters. The process feature construction module is used to construct the process state features of the spiral conveyor channel based on the operating parameters. The process state features include at least: load distribution parameters determined according to the driving load parameters, load gradient parameters calculated based on the load distribution parameters, load evolution rate parameters calculated based on the time change of the load gradient parameters, and conveying stability parameters calculated based on the energy consumption parameters and the soil discharge continuity parameters. Equivalent conveying capacity calculation module: used to calculate the equivalent conveying capacity parameters characterizing the conveying capacity of the screw conveyor system based on the load distribution parameters, conveying stability parameters, and energy consumption parameters; The conveying status and capacity determination module is used to determine the conveying status category and its development trend of slag and soil based on the load gradient parameter and load evolution rate parameter, determine the soil silo pressure maintenance requirement parameter based on the soil silo pressure change rate, and determine the degree of mismatch between conveying capacity and pressure maintenance requirement based on the relationship between the equivalent conveying capacity parameter and the soil silo pressure maintenance requirement parameter. Dynamic control window generation module: used to generate a multi-parameter dynamic collaborative control window based on the soil transport status category, development trend and mismatch degree. The control window includes at least: allowable range of soil discharge capacity, allowable range of tunneling advance and soil improvement adjustment range. Deviation determination module: used to determine the system operation deviation and its degree based on the relationship between the current operating parameters and the multi-parameter dynamic collaborative control window; Closed-loop reconfiguration control module: used to perform the following operations when there is an operational deviation: based on the load gradient parameters and load evolution rate parameters, determine the abnormal section and its influence range; based on the influence range, establish the coupling relationship between each section in the spiral conveying channel; based on the degree of mismatch, coordinately adjust the soil discharge capacity parameters, soil chamber pressure maintenance parameters and tunneling advancement parameters; and gradually adjust the soil conveying state according to the preset adjustment sequence so that the system evolves from an abnormal state to a stable conveying state. Steady-state maintenance module: used to maintain the coordinated and stable operation of the slag conveying and soil pressure when the operating parameters are within the multi-parameter dynamic coordinated control window.

2. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 1, characterized in that: The equivalent conveying capacity calculation module is used to calculate the equivalent conveying capacity parameters based on the operating parameters, including segment division and load feature extraction: based on the drive load parameters of each axial position of the screw conveyor, a load distribution sequence along the path is constructed, and the screw conveying channel is divided into at least two axial segments, and the segment load feature parameters corresponding to each axial segment are determined respectively. Load evolution characteristics construction: Based on the load characteristic parameters of each section, determine the section load gradient parameters and section load non-uniformity coefficient of each axial section, and determine the section load evolution rate parameters based on the changes of the section load gradient parameters within a preset time window. Construction of transport energy efficiency and continuity characteristics: Based on the energy consumption parameters and the change in soil discharge volume, the unit volume transport energy consumption parameters are determined, and based on the soil discharge continuity parameters, the soil discharge continuity coefficient and soil discharge discontinuity frequency parameters are determined. Construction of resistance characterization parameters: Based on the section load gradient parameters, section load non-uniformity coefficient and section load evolution rate parameters, resistance characterization parameters for characterizing the resistance of slag and soil transportation are constructed. Construction of capacity characterization parameters: Based on the unit volume transport energy consumption parameter, soil discharge continuity coefficient and soil discharge discontinuity frequency parameter, capacity characterization parameters for characterizing the capacity of slag transport are constructed. Adaptive determination of segment weights: A weight coefficient is set for each axial segment, and the weight coefficient is adaptively adjusted based on the segment load gradient parameter and the segment load evolution rate parameter; Coupled calculation and capacity mapping: Based on the capacity characterization parameters and resistance characterization parameters of each axial segment, a weighted coupled calculation is performed, and the equivalent transport capacity parameter is determined based on the preset capacity-resistance mapping relationship; Wherein: the capacity-resistance mapping relationship is used to characterize the correspondence between the capacity and resistance of the slag transport, the weighting coefficient is used to characterize the contribution of different axial sections to the overall transport capacity, and the adaptive adjustment is: adjusting the weighting coefficient by increasing or decreasing based on the direction and rate of change of the section load gradient.

3. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 2, characterized in that: The segment coupling control module is used to determine abnormal segments and their impact range based on the operating parameters, including multi-scale segment division: constructing a load distribution sequence along the axial direction based on the drive load parameters of each axial position of the screw conveyor, and performing adaptive multi-scale segment division based on the load distribution sequence, wherein the segment division granularity is adjusted according to the change amplitude and change density of the load distribution sequence to obtain segment division results of at least two different scales. Candidate anomalous segment identification: Under the segment division results at each scale, the segment load gradient parameters of each segment are determined respectively, and local abrupt change points in the load gradient distribution are identified based on the segment load gradient parameters. When the load gradient change amplitude of the corresponding segment is greater than the preset gradient abrupt change threshold, the segment is identified as a candidate anomalous segment. Abnormal segment determination: Within a preset time window, the segment load gradient parameters of the candidate abnormal segment are analyzed over time. When the segment load gradient parameters show a monotonic change trend in multiple consecutive sampling periods and the rate of change is greater than a preset evolution rate threshold, the candidate abnormal segment is determined to be an abnormal segment. Expansion of the abnormal impact range: Taking the abnormal section as the center, under the segment division results of different scales, the segment is expanded upstream and downstream along the axial direction of the spiral conveying channel, respectively. The segment load gradient parameters, segment load non-uniformity coefficient and soil discharge continuity parameters of adjacent segments are jointly judged. When the preset segment association conditions are met, the adjacent segments are included in the abnormal impact range. Abnormal propagation boundary determination: During the segment expansion process, the attenuation characteristics of the segment load gradient parameter and the segment load evolution rate parameter of adjacent segments are analyzed. When the change amplitude of the segment load gradient parameter is lower than the preset attenuation threshold and the segment load evolution rate parameter is lower than the preset stability threshold, the corresponding segment is determined as the abnormal propagation boundary segment, and the segment expansion along this direction is stopped. Determination of abnormal expansion characteristics: Based on the spatial distribution of the segment load evolution rate parameters and the segment load non-uniformity coefficient of each segment within the scope of the abnormal influence, the main expansion direction and expansion speed of the abnormal segment are determined. Anomaly propagation path construction and trend prediction: Based on the main expansion direction, expansion speed and anomaly propagation boundary segment, anomaly propagation path is constructed, and based on the anomaly propagation path and segment load evolution rate parameters, the development trend of abnormal soil and waste transportation is predicted.

4. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 3, characterized in that: The dynamic control window generation module is used to generate the multi-parameter dynamic collaborative control window based on the operating parameters, including the division of operating stages: based on the equivalent transport capacity parameter and the earthen silo pressure change rate, the earthen silo pressure maintenance requirement parameter is determined, and based on the difference between the equivalent transport capacity parameter and the earthen silo pressure maintenance requirement parameter and its rate of change, the transport capacity-pressure requirement deviation parameter is determined, and based on the magnitude, direction of change and rate of change of the deviation parameter, the current operating state is divided into stages, the stages including at least a stable stage, a transition adjustment stage and an abnormal control stage; Determination of target range for soil discharge capacity: Under different operating stages, the corresponding stage control model is called respectively. Based on the conveying capacity-pressure demand deviation parameter and its rate of change, the target range for soil discharge capacity is determined. In the transition adjustment stage, the soil discharge capacity is smoothly adjusted based on the continuous mapping function. In the abnormal control stage, the soil discharge capacity is restricted and adjusted based on the amplitude limit constraint function. Construction of tunneling propulsion parameter constraint range: Based on the target range of soil removal capacity and the rate of change of soil pressure, the allowable fluctuation range of soil pressure is determined, and based on the coupling constraint relationship between the target range of soil removal capacity and the allowable fluctuation range of soil pressure, the allowable adjustment range of tunneling propulsion parameters is determined through a multivariate constraint mapping function; Spatial constraint correction: Based on the main expansion direction, expansion speed and abnormal propagation boundary section of the abnormal section, a spatial constraint correction function is constructed to make directional corrections to the target range of the soil removal capacity and the allowable adjustment range of the tunneling advancement parameters. When approaching the abnormal propagation boundary section, the adjustment range is dynamically adjusted through a progressive recovery function. Risk constraint correction: Based on the dumping discontinuity frequency parameter, transport stability parameter and risk index, a risk constraint function is constructed, and the boundary correction is performed on the target range of dumping capacity, the allowable fluctuation range of earth chamber pressure and the allowable adjustment range of tunneling advance parameters based on the risk constraint function; Historical working condition correction: Based on historical tunneling data and current operating status parameters, a historical working condition matching model is constructed, and the target range of soil removal capacity and the allowable adjustment range of tunneling advancement parameters are corrected based on the historical working condition matching model; Construction of collaborative control domain: The target range of soil dumping capacity, the allowable fluctuation range of soil pressure, the allowable adjustment range of tunneling propulsion parameters, and the adjustment range of slag improvement are uniformly constrained and mapped to construct a multi-parameter collaborative control domain; Dynamic scheduling and control window generation: Within the multi-parameter collaborative control domain, based on a preset multi-objective optimization scheduling model, each control parameter is scheduled and dynamically updated in a time sequence, so that the operating parameters gradually tend to a stable delivery state, thereby generating the multi-parameter dynamic collaborative control window.

5. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 4, characterized in that: The closed-loop reconfiguration control module is used to perform execution control and feedback updates on the multi-parameter dynamic collaborative control window, including execution sequence determination: based on the allowable adjustment range of each control parameter in the multi-parameter dynamic collaborative control window, a set of candidate adjustment quantities for each execution object is determined, and based on the delivery capacity-pressure demand deviation parameter and the main expansion direction of the abnormal section, the set of candidate adjustment quantities is prioritized to obtain the execution sequence; Conflict resolution processing: Based on the execution sequence, conflict determination is performed on the candidate adjustment quantities of different execution objects. When the candidate adjustment quantities of different execution objects have opposite effects on the same constraint variable, the candidate adjustment quantities are corrected based on the preset conflict resolution rules to obtain a consistent adjustment quantity. Phased adjustment and time-series decoupling control: Based on the execution sequence, each execution object is adjusted in stages, and time-series decoupling control is implemented on the parameters related to soil dumping capacity and the tunneling propulsion parameters. In the soil dumping capacity adjustment stage, the tunneling propulsion parameters are subject to maintenance or gradual change constraints, and in the propulsion adjustment stage, the soil dumping capacity parameters are subject to compensation constraints. Dynamic constraint correction: During the adjustment process, based on the soil pressure change rate and load evolution rate parameters, the change rate limit and hysteresis compensation processing are applied to the adjustment amount of each execution object. The hysteresis compensation processing is used to reduce the impact of the response lag of the execution object on the soil pressure. Security degradation control: When the expansion speed of the abnormal segment exceeds the preset threshold or the risk indicator reaches the preset threshold, security degradation control is triggered to apply a limiting constraint to the adjustment amount of the execution object and narrow the value range of the multi-parameter dynamic collaborative control window. Predictive compensation correction: In each adjustment cycle, based on the expansion rate of the abnormal section and the changing trend of the conveying capacity-pressure demand deviation parameter, the target adjustment amount for the next adjustment cycle is predicted and compensated. Parameter update: After each adjustment cycle, the operating parameters are reacquired, and the equivalent conveying capacity parameters and the conveying capacity-pressure demand deviation parameters are re-determined based on the updated operating parameters. Closed-loop convergence control: Based on the updated delivery capacity-pressure demand deviation parameter, compare it with the preset convergence condition: when the convergence condition is not met, adaptively adjust the adjustment step size based on the change amplitude and change rate of the deviation parameter, and progressively update the target adjustment amount, and enter the next adjustment cycle; when the convergence condition is met, maintain the current operating state and enter the steady-state maintenance mode.

6. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 5, characterized in that: The system also includes a risk protection and control module, which is used for risk identification and safety control, including the construction of multiple risk indicators: based on the operating parameters and their changing relationships obtained by the multi-segment data acquisition module, the gushing risk indicator, the soil chamber pressure drop risk indicator and the transport blockage risk indicator are determined respectively, and multiple risk coupling risk parameters are constructed based on the correlation relationship between each risk indicator. Dynamic risk threshold determination: Based on the statistical distribution characteristics of the operating parameters and historical tunneling data, the risk thresholds corresponding to each risk indicator are updated to obtain dynamic risk thresholds, and the risk development trend parameters are determined based on the relationship between the multi-risk coupled risk parameters and the dynamic risk thresholds. Risk level determination: Based on the multi-risk coupling risk parameters and the risk development trend parameters, a risk level determination is made to obtain the risk level corresponding to the current operating status. The risk level includes at least normal status, early warning status, risk status and dangerous status. Risk constraint embedding and window shrinking: In the early warning state or risk state, the multi-risk coupled risk parameters are embedded as constraint variables into the multi-parameter dynamic collaborative control window, and based on the risk constraint mapping relationship, the range of the target range of soil dumping capacity, the allowable fluctuation range of soil pressure and the allowable adjustment range of tunneling advancement parameters are shrunk, and the shrinkage range is adjusted based on the risk development trend parameters; Graded safety degradation control: Under the dangerous state, graded safety degradation control is triggered to adjust the system according to the preset degradation sequence, including reducing the upper limit of soil dumping capacity, limiting tunneling propulsion parameters, and adjusting soil improvement parameters, and applying coordination constraints and change rate limits to each execution object; Dynamic correction during the degradation process: During the implementation of the security degradation control, the degradation control strategy is adjusted based on the multi-risk coupling risk parameters and the risk development trend parameters, and further degradation control is executed when the risk is detected to be further increased. Phased recovery control: When the multi-risk coupling risk parameters are reduced to below the corresponding recovery threshold, phased rollback and recovery scheduling control is executed. The constraints on soil dumping capacity, tunneling propulsion parameters, and slag improvement parameters are relaxed in sequence, and the multi-parameter dynamic collaborative control window is gradually restored, so that the system switches from the safety control mode to the collaborative regulation and control operation state.

7. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 6, characterized in that: The system also includes a central control unit, an edge processing unit, and a data communication module. The central control unit is connected to a multi-segment data acquisition module, a process feature construction module, an equivalent transport capacity calculation module, a transport status and capacity determination module, a dynamic control window generation module, a closed-loop reconfiguration and control module, and a risk protection control module, respectively, and is used to execute global scheduling control and generate multi-parameter dynamic collaborative control windows. Edge processing unit: The edge processing unit is located between the multi-segment data acquisition module and the central control unit. It is used to preprocess the acquired operating parameters, including data filtering, abnormal data removal and preliminary feature extraction, and to execute local control response when the rate of change of operating parameters exceeds a preset threshold. Data communication module: The data communication module includes a data transmission channel and a control command channel. The data transmission channel is used to transmit operating parameters from the multi-segment data acquisition module and the edge processing unit to the central control unit. The control command channel is used to send control commands generated by the central control unit to each execution object. Dual-plane architecture: The central control unit is used to construct a dual-plane architecture of a data processing plane and a control execution plane. The data processing plane is used to perform features along the route, equivalent transport capacity calculation and abnormal section identification. The control execution plane is used to perform dynamic control window generation, closed-loop reconstruction regulation and risk protection control, and realizes decoupled interaction between data processing and control execution through the inter-plane interface. Time synchronization and causal relationship: The central control unit is used to perform time synchronization processing on the data output by each module, and to establish the correspondence between operating parameters and control commands based on timestamps; Control command consistency processing: The central control unit is used to process control commands from different modules for consistency. When there is a conflict between control commands output by different modules, the control commands are corrected based on preset priority rules and risk levels, and the corrected control commands are sent to the execution object. Multi-level data caching: The central control unit is used to establish multi-level data caching areas among the modules, store the operating parameters and calculation results at different time scales in layers, and provide historical data to the equivalent delivery capacity calculation module and risk protection control module according to the calling requirements; Abnormal operating condition fault tolerance control: When the data communication module experiences a communication abnormality or the central control unit experiences an operational abnormality, the edge processing unit executes fault tolerance control according to a preset degradation takeover strategy, including local control, amplitude limiting control and safe operation control.

8. The shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 7, characterized in that: The execution objects include a screw conveyor drive unit, a tunneling propulsion unit, and a slag improvement unit. The screw conveyor drive unit includes a frequency converter drive unit, a multi-stage speed adjustment unit, and a speed change rate constraint unit. The frequency converter drive unit is used to adjust the speed of the screw conveyor according to the control commands output by the central control unit. The multi-stage speed adjustment unit is used to select the corresponding speed adjustment mode at different operating stages. The speed change rate constraint unit is used to limit the speed change rate. Tunneling propulsion device: The tunneling propulsion device includes a propulsion cylinder, a propulsion pressure regulating unit, and a propulsion speed coordinating unit. The propulsion pressure regulating unit is used to regulate the output pressure of the propulsion cylinder, and the propulsion speed coordinating unit is used to adjust the propulsion speed based on the changes in the rotational speed of the screw conveyor. Slag Soil Improvement Device: The slag soil improvement device includes an amendment delivery unit, a flow regulation unit, and an improvement parameter feedback unit. The flow regulation unit is used to regulate the injection flow rate of the amendment, and the improvement parameter feedback unit is used to obtain the injection status of the amendment and feed it back to the central control unit. Control interface module: The central control unit is connected to each execution device through the control interface module. The control interface module includes a signal conversion unit, a signal isolation unit, a redundant communication unit, and a dual-channel drive unit. The dual-channel drive unit is used to provide a main drive channel and a backup drive channel for key execution objects. Cooperative constraint processing: The central control unit is used to construct cooperative constraint relationships between execution objects, to describe the coupling relationship between the screw conveyor speed, propulsion parameters and improvement parameters, and to perform constraint correction on the control commands of each execution object based on the cooperative constraint relationships; Feedback fusion and consistency processing: The control interface module is used to receive feedback signals from each execution device, and to fuse the feedback signals through the multi-source feedback fusion processing unit. At the same time, the consistency verification unit verifies the fusion result to obtain the execution status parameters. Control compensation correction: The central control unit is used to compensate and correct the control command based on the deviation between the execution status parameters and the target control command, so as to reduce the impact of execution lag, nonlinear response and load changes on the control effect; Fault isolation and redundancy switching: When the control interface module detects an abnormality in any execution device or communication channel, the abnormal execution path is isolated by the fault isolation unit, and the dual-channel drive unit switches to the backup drive channel, while the corresponding execution object is subjected to amplitude limiting control.

9. A shield tunneling excavation soil conveying and soil chamber pressure coordinated control system according to claim 8, characterized in that: The equivalent conveying capacity calculation module is used to determine the equivalent conveying capacity parameters based on the operating parameters, including segment statistical feature determination: based on the segment load parameters of each axial segment of the screw conveyor, the average load value and load dispersion parameters of each segment are determined respectively, and the segment load non-uniformity coefficient is determined based on the average load value and load dispersion parameters. At the same time, the segment weight coefficient is determined according to the position of each segment in the screw conveying channel. Determination of temporal evolution characteristics: Based on the changes in the load parameters of the segment within a preset time window, the load change rate and load fluctuation intensity parameters of the segment are determined, and a temporal memory item for characterizing the temporal evolution characteristics is constructed based on the load change characteristics of the segment. Energy efficiency and continuity characteristics determination: Based on the energy consumption parameters and the change in soil discharge volume, the energy consumption parameters for transporting per unit volume are determined, and based on the change in the soil discharge continuity parameters within a preset time window, the soil discharge interruption frequency parameters are determined. Construction of capacity and resistance characterization parameters: Based on the section load non-uniformity coefficient, section weight coefficient and time sequence memory term, resistance characterization parameters for characterizing transport resistance are constructed, and based on the unit volume transport energy consumption parameter and soil discharge intermittent frequency parameter, capacity characterization parameters for characterizing transport capacity are constructed. Coupling feature construction: Based on the transport state category, the resistance characterization parameters and capacity characterization parameters are subjected to state discrimination processing, and the confidence of each parameter is combined to weight them to obtain the capacity-resistance coupling feature quantity; Equivalent transport capacity determination: Based on the capacity-resistance coupling characteristic quantity and the preset mapping relationship, the equivalent transport capacity parameter is determined, and the mapping relationship is updated based on historical operating data and current calculation results; Consistency constraint correction: The equivalent transport capacity parameters are subjected to consistency constraint correction to make them consistent with the change rate of soil pressure and the change trend of soil discharge capacity.

10. A method for coordinated control of excavated soil transportation and soil chamber pressure during shield tunneling, used in the system described in any one of claims 1-9, characterized in that, Includes the following steps: S1: Operational parameter acquisition and preprocessing: Acquire soil pressure, drive load parameters of each axial section of the screw conveyor, energy consumption parameters and soil discharge continuity parameters during shield tunneling construction, and perform filtering, abnormal data removal and time synchronization processing on the operational parameters. S2: Construction of load characteristics along the conveyor: Based on the driving load parameters, the load distribution along the conveyor is constructed, and the non-uniformity coefficient of the section load, the rate of load change and the intensity of load fluctuation are determined. Based on the load change characteristics, a time-series memory term is constructed to characterize the time evolution characteristics. S3: Energy efficiency and continuity characteristics construction: Based on the energy consumption parameters and the change in soil discharge volume, determine the energy consumption parameters for transporting per unit volume, and determine the soil discharge interruption frequency parameters based on the soil discharge continuity parameters. S4: Determination of equivalent conveying capacity: Based on the section load non-uniformity coefficient, time sequence memory term, unit volume conveying energy consumption parameter and soil discharge intermittent frequency parameter, construct conveying resistance characterization parameters and conveying capacity characterization parameters, and determine the equivalent conveying capacity parameter based on the mapping relationship between the resistance characterization parameters and the capacity characterization parameters. S5: Operation status determination: Based on the relationship between the equivalent conveying capacity parameter and the earth chamber pressure change rate, determine the conveying capacity-pressure demand deviation parameter, and based on the magnitude, direction of change and rate of change of the deviation parameter, divide the current operation status into stages, the stages including at least a stable stage, a transitional adjustment stage and an abnormal control stage. S6: Control window generation: Based on the running status and abnormal segment identification results, a multi-parameter dynamic collaborative control window is generated, and the control parameters are limited within the range of the control window. At the same time, the control window is adjusted according to the expansion direction of the abnormal segment. S7: Coordinated Adjustment Control: Under different operating stages, based on the preset scheduling model, the speed of the screw conveyor, the tunneling propulsion parameters and the soil improvement parameters are coordinated to take into account the stability of the soil chamber pressure, the conveying efficiency and the energy consumption level. S8: Temporal decoupling and constraint control: During the adjustment process, temporal decoupling control is implemented for the adjustment of soil dumping capacity and the adjustment of tunneling advance, and change rate constraints and coordination constraints are applied to each control parameter; S9: Predictive compensation adjustment: In each adjustment cycle, based on the changing trend of the deviation parameter and the expansion rate of the abnormal section, the control parameter of the next adjustment cycle is compensated and corrected. S10: Closed-loop update and convergence control: During the adjustment process, the updated operating parameters are obtained, and the equivalent transport capacity parameters and the deviation parameters are re-determined based on the updated operating parameters to perform closed-loop adjustment; and the adjustment step size is adjusted based on the change amplitude and change rate of the deviation parameters so that the system gradually tends to a stable operating state. S11: Risk Identification and Classification Control: Based on the gushing risk index, the soil chamber pressure drop risk index and the transport blockage risk index, a multi-risk coupled risk parameter is constructed, and the system operating status is classified based on the risk parameter. Under the risk status, the range of the multi-parameter dynamic collaborative control window is adjusted or safety degradation control is executed. S12: Risk recovery control: During the risk control process, the control strategy is adjusted based on the risk development trend, and phased recovery control is executed when the risk parameters are reduced to the preset recovery conditions, so that the system operation status is gradually restored to the coordinated control operation status.

Images