A self-adaptive grease injection construction method and system for shield tail sealing grease

Abstract

The present application relates to a kind of shield tail sealing grease self-adapting injection construction method and system, wherein the method, comprising: the state perception parameter of construction area is synchronously collected with set frequency;According to state perception parameter, target pressure, circumferential pressure unevenness and demand injection rate are calculated;According to target pressure, circumferential pressure unevenness and demand injection rate, optimal injection rate and the distribution coefficient of each injection point are solved;Control servo motor and variable displacement pump of shield tail sealing system according to optimal injection rate operation, each electromagnetic proportional distribution valve is adjusted opening according to distribution coefficient.The present application solves optimal injection rate and the distribution coefficient of each injection point by target pressure, circumferential pressure unevenness, demand injection rate multiple constraint condition, can realize shield tail circumferential each sealing point position balanced grease supply, prevent local lack of fat dry grinding or excessive injection flush seal.

Description

Technical Field

[0001] This invention relates to the field of shield tunnel construction technology, specifically to an adaptive grease injection method and system for shield tail sealing grease. Background Technology

[0002] The shield tail sealing system is the core safety barrier of the tunnel boring machine. By continuously injecting special sealing grease into the sealing cavity between the shield tail shell and the tunnel segments, a dynamic flexible sealing layer is formed to resist external water and soil pressure and prevent groundwater, mud and additives from seeping into the tunnel. The grease injection technology commonly used in shield construction at home and abroad has the following fundamental defects: (1) Operation depends on manual experience and has a low degree of automation: Traditional grease injection systems are mostly open-loop modes of "timed and quantitative" or "pressured start and stop". The grease injection interval and grease injection amount need to be set by the operator based on experience. When encountering sudden changes in strata, the parameters must be manually adjusted, and the response time is long (usually 15~30 minutes), which can easily cause insufficient or excessive grease injection. (2) The coupling interference of multiple factors is serious and lacks self-adaptive ability: Even if the existing system adopts closed-loop PID control, its control parameters (proportional, integral and derivative gain) are fixed values, which cannot adapt to dynamic interferences such as soil pressure fluctuations, grease viscosity drift caused by changes in ambient temperature, and volume changes caused by wear of the sealing cavity during the tunneling process. Actual engineering data shows that when the grease temperature rises from 20℃ to 40℃, the viscosity decreases by about 40%, and the sealing cavity pressure decreases by 0.2~0.4MPa under the same timed grease injection volume, significantly increasing the risk of leakage. The existing system is powerless to address this. (3) Blind spots in circumferential sealing monitoring lead to local failure: Most equipment can only indirectly judge the sealing status through pump outlet pressure or single-point sealing cavity pressure, and cannot sense the real-time gradient of the circumferential pressure distribution. Especially at the 6 o'clock position (bottom of the pipe segment), due to grease settling and its own weight, the grease injection volume is seriously insufficient, leading to frequent grout leakage accidents. (4) Lack of fault self-diagnosis and self-healing capabilities: The existing system lacks real-time diagnostic functions for faults such as grease pipeline blockage, grease deterioration, and pump wear. It often only alarms when the seal fails, resulting in long unplanned downtime and seriously affecting the construction progress.

[0003] Therefore, there is an urgent need to develop a shield tail sealing grease injection equipment and construction method that can achieve fully automated operation, dynamic adaptive adjustment, precise circumferential grease injection, and self-healing ability. This is a pressing need to improve the safety and economy of shield tunneling construction. Summary of the Invention

[0004] To address the aforementioned problems, the purpose of this invention is to provide an adaptive grease injection method and system for shield tail sealing grease.

[0005] A method for adaptive grease injection for shield tail sealing includes:

[0006] Step 1: Synchronously collect status sensing parameters of the construction area at a set frequency; the status sensing parameters include: construction area pressure, grease temperature, soil chamber pressure, mass change rate, instantaneous flow rate and vibration signal;

[0007] Step 2: Calculate the target pressure, circumferential pressure non-uniformity, and required grease injection rate based on the status sensing parameters;

[0008] Step 3: Based on the target pressure, circumferential pressure unevenness, and required grease injection rate, calculate the optimal grease injection rate and the distribution coefficient of each grease injection point;

[0009] Step 4: Control the servo motor and variable piston pump of the shield tail sealing system to operate at the optimal grease injection rate, and adjust the opening of each electromagnetic proportional distribution valve according to the distribution coefficient.

[0010] Preferably, in step 2, the formula for calculating the required grease injection rate is:

[0011]

[0012] Where K1, K2, and K3 are adaptive gain coefficients; For the required grease injection rate; P target For target pressure; P soil For earth chamber pressure; For time; Circumferential pressure; Q0 is the baseline grease injection rate; µ(T) is the temperature-viscosity compensation function. α is the viscosity temperature coefficient, T0 is the reference temperature; P target =P soil +ΔP safe ΔP safe For safety margin, the large-diameter high-water-pressure shield tunnel ΔP safe Automatically adjusts to 0.6~0.8MPa; under low-temperature conditions (T<0℃), synchronously links the oil tank constant temperature heating device, amplifies K3 to compensate for high viscosity flow resistance, and extends the duration of grease injection.

[0013] Preferably, in step 2, the adaptive gain coefficients K1, K2, and K3 are determined using a dual-mode approach of offline calibration and online incremental iteration. Offline, static values ​​are assigned based on geological surveys and oil rheology tests. During the tunneling process, small corrections are made per ring using the gradient descent method, with each correction not exceeding 10% of the initial value. Corrections are only performed when the performance index of that ring is better than the historical average; otherwise, the correction is skipped and the abnormal event is recorded to avoid abnormal data contaminating the gain coefficients. All fault judgment thresholds are dynamically and adaptively corrected based on rolling historical working conditions every 30 rings. The pressure is relaxed to 0.35–0.45 MPa for soft soil strata and tightened to 0.20–0.28 MPa for gravel strata. The earth pressure change warning threshold is adjusted upward within the range of 0.08–0.12 MPa / s for every 10m increase in burial depth.

[0014] Preferably, in step 3, a multi-objective weighted loss function is constructed using model predictive control, which includes a pressure tracking error term, a grease consumption term, and a circumferential distribution balance term. Triple hard constraints are set: upper and lower limits for flow rate, normalization of distribution coefficients, and minimum safe sealing pressure. The weighting coefficients of the loss function are adaptively tuned according to formation water pressure and construction cost requirements. For high-water-pressure formations, the pressure weight is increased; for cost-control items, the grease consumption weight is increased. Simultaneously, shield machine adaptation rules are built-in, and after inputting the shield diameter, number of sealing channels, and number of grease injection points, the benchmark distribution coefficient and benchmark grease injection rate are automatically corrected. The proportion of basic flow at the bottom point of large shield tunnels with a diameter greater than 8m increases by 10% to 20%.

[0015] Preferably, in step 3, a model predictive control algorithm is used to predict the pressure evolution over the next 5 seconds, and the optimal grease injection rate sequence that minimizes the weighted sum of the squared pressure tracking error and grease consumption is solved, with a control period of 100 ms; the complete mathematical expression of the multi-objective loss function is as follows: In the formula =5s prediction time domain, To control the time domain, Pressure safety weight, Fat consumption weight Circumferential equilibrium weights; Predicting pressure at the grease injection point, For single-point allocation coefficients, The average distribution coefficient; constraints: .

[0016] Preferably, in step 3, the prediction model in the model prediction control algorithm is established based on the circumferential pressure sequence, soil chamber pressure sequence, grease injection rate sequence, and grease temperature sequence in the historical tunneling data of the tunnel boring machine, through a system identification method.

[0017] This invention also provides an adaptive grease injection system for tail shield sealing, comprising:

[0018] The parameter acquisition module is used to synchronously acquire the status sensing parameters of the construction area at a set frequency; the status sensing parameters include: construction area pressure, grease temperature, soil chamber pressure, mass change rate, instantaneous flow rate and vibration signal.

[0019] The parameter calculation module is used to calculate the target pressure, circumferential pressure unevenness, and required grease injection rate based on the state-sensing parameters; it includes built-in offline calibration programs for K1, K2, and K3, a dynamic threshold adaptive correction unit, viscosity compensation linkage logic across the high and low temperature range, and a shield machine model parameter adaptation submodule.

[0020] The working parameter calculation module is used to solve for the optimal grease injection rate and the distribution coefficient of each grease injection point based on the target pressure, circumferential pressure non-uniformity and required grease injection rate; it integrates a complete MPC multi-objective loss function solver, weighted coefficient adaptive tuning unit and differentiated distribution coefficient correction algorithm, and supports multi-constraint boundary condition determination.

[0021] The operation control module controls the servo motor and variable piston pump of the shield tail sealing system to operate at the optimal grease injection rate, and adjusts the opening of each electromagnetic proportional distribution valve according to the distribution coefficient; it has a built-in fault classification and scheduling unit with three levels of fault priority: first-level faults (main pump failure, main pipe overpressure) are handled first, second-level faults (single-point blockage, local leakage) and third-level faults (temperature drift, slight pressure unevenness) are cached in the task queue, and self-healing actions are executed sequentially after the high-level faults are eliminated; it is equipped with dual hardware and software safety interlocks: two-stage pressure relief of electric relief valve and mechanical safety valve, hardware limiting of flow output, and automatic recovery function of power failure parameter caching upon power-on.

[0022] Preferably, in the parameter calculation module, the formula for calculating the required grease injection rate is:

[0023]

[0024] Where K1, K2, and K3 are adaptive gain coefficients; For the required grease injection rate; P target For target pressure; P soil For earth chamber pressure; For time; Circumferential pressure; Q0 is the baseline grease injection rate; µ(T) is the temperature-viscosity compensation function. α is the viscosity temperature coefficient, T0 is the reference temperature; P target =P soil +ΔP safe ΔP safe For safety margin.

[0025] Preferably, in the working parameter calculation module, a model predictive control algorithm is used to predict the pressure evolution over the next 5 seconds, and to solve for the optimal grease injection rate sequence that minimizes the weighted sum of the pressure tracking error squared and grease consumption, with a control period of 100 ms.

[0026] Preferably, in the working parameter calculation module, the prediction model in the model prediction control algorithm is established based on the circumferential pressure sequence, soil chamber pressure sequence, grease injection rate sequence, and grease temperature sequence in the historical tunneling data of the tunnel boring machine, through a system identification method.

[0027] The present invention also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, characterized in that the computer program, when executed by the processor, implements the steps in the above-described adaptive grease injection method for shield tail sealing.

[0028] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps in the above-described adaptive grease injection method for shield tail sealing.

[0029] This invention relates to an adaptive grease injection method for shield tail sealing. Compared with existing technologies, this invention solves for the optimal grease injection rate and the distribution coefficient of each grease injection point by considering multiple constraints such as target pressure, circumferential pressure unevenness, and required grease injection rate. This achieves balanced grease supply to all sealing points in the circumferential direction of the shield tail, preventing localized dry grinding due to insufficient grease or excessive grease injection that erodes the seal. Simultaneously, it addresses logical and engineering deficiencies in existing schemes, such as lack of initial parameter data, fixed thresholds, conflicting multiple actions, poor versatility, and data insecurity, through gain standardization calibration, dynamic fault thresholds, multi-fault hierarchical self-healing, cross-shield machine adaptive design, complete MPC mathematical constraints, low-temperature full-condition compensation, and federated learning privacy aggregation mechanism.

[0030] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 A schematic diagram of an adaptive grease injection method for shield tail sealing grease provided by the present invention;

[0033] Figure 2 This invention provides a schematic diagram of the adaptive grease injection construction principle for shield tail sealing grease. Detailed Implementation

[0034] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0035] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0036] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0037] Please see Figure 1 An adaptive grease injection method for shield tail sealing grease, comprising:

[0038] Step 1: Synchronously collect status sensing parameters of the construction area at a set frequency; the status sensing parameters include: construction area pressure, grease temperature, soil chamber pressure, mass change rate, instantaneous flow rate and vibration signal;

[0039] Step 2: Calculate the target pressure, circumferential pressure non-uniformity, and required grease injection rate based on the status sensing parameters;

[0040] Step 3: Based on the target pressure, circumferential pressure unevenness, and required grease injection rate, calculate the optimal grease injection rate and the distribution coefficient of each grease injection point;

[0041] Step 4: Control the servo motor and variable piston pump of the shield tail sealing system to operate at the optimal grease injection rate, and adjust the opening of each electromagnetic proportional distribution valve according to the distribution coefficient.

[0042] The working principle of the present invention will be explained below with reference to specific embodiments:

[0043] Step S1: System Initialization and Self-Test

[0044] After the equipment is powered on, it performs a self-test on the tail shield sealing system and displays the results; it reads the historical tunneling parameters stored in the memory and the cloud-based pre-trained model, and loads the initial control parameters. It then starts the machine adaptation program, inputs the shield diameter, number of sealing channels, and number of grease injection points, and automatically corrects Q0, the initial allocation coefficient, and the safety margin ΔP. safe Perform offline calibration loading of K1, K2, and K3, and match the initial value of the basic gain according to the formation type.

[0045] Step S2: Real-time acquisition of multiple parameters

[0046] After the tunnel boring machine begins excavation, the sensing unit synchronously collects data at a frequency of 100 Hz: the circumferential pressure vector P measured by the pressure sensor. i (t) (i=1~N, N≥8), grease temperature T(t) measured by temperature sensor, and grease pressure P measured by grease chamber pressure transmitter. soil (t), the rate of change of mass measured by the weighing sensor dM / dt, the instantaneous flow rate Q(t) measured by the flow sensor, and the vibration signal V(t) measured by the vibration sensor.

[0047] Step S3: The microprocessor executes an adaptive decision loop every 100 ms.

[0048] Calculate the target pressure P target(t) =P soil (t) +ΔP safe , where ΔP safe For safety margin, the default value is 0.5 MPa;

[0049] Calculate the circumferential pressure non-uniformity ΔPmax = max(Pi) - min(Pi). If ΔPmax > 0.3MPa, it is determined that the pressure distribution is uneven and circumferential distribution optimization is initiated.

[0050] Calculate the required grease injection rate:

[0051]

[0052] Where K1, K2, and K3 are adaptive gain coefficients, determined using a dual-mode approach of offline calibration and online incremental iteration. Offline, static values ​​are assigned based on geological surveys and grease rheology tests. During tunneling, small corrections are made per ring using the gradient descent method, with each correction not exceeding 10% of the initial value. Corrections are only performed when the performance indicators of that ring are better than the historical average; otherwise, the correction is skipped and the abnormal event is recorded. µ(T) is the temperature-viscosity compensation function, μ0 is the grease viscosity at the reference temperature, and Q0 is the baseline grease injection rate. The temperature-viscosity compensation function uses: (α is the viscosity temperature coefficient, T0 is the reference temperature); when T < 0℃, the oil tank heating is started simultaneously to increase K3 to compensate for high viscosity resistance; when T > 50℃, the target pressure is automatically reduced by 0.1MPa, and the grease injection frequency is increased simultaneously.

[0053] Model predictive control optimization: Taking the pressure evolution over the next 5 seconds as the prediction target, the following multi-objective weighted loss function is constructed: In the formula =5s prediction time domain, To control the time domain, Pressure safety weight, Fat consumption weight Circumferential equilibrium weights; Predicting pressure at the grease injection point, For single-point allocation coefficients, The average distribution coefficient is used; constraints include upper and lower limits for flow rate. Find the optimal grease injection rate Q that minimizes the weighted sum of the squared pressure tracking error and grease consumption. opt (t) and the distribution coefficient β of each injection point i (t). The prediction model is established based on the circumferential pressure sequence, soil chamber pressure sequence, grease injection rate sequence, and grease temperature sequence from the historical tunneling data of the tunnel boring machine, and is established through a system identification method.

[0054] Step S4: Precise liposuction execution

[0055] Servo motor drives variable displacement piston pump according to Q opt (t) Operation; each electromagnetic proportional distribution valve operates according to the distribution coefficient β i (t) Adjust the opening degree; the electromagnetic proportional relief valve adjusts the upper limit of the system pressure; simultaneously, the weighing sensor provides real-time feedback on the actual injection rate, which is consistent with Q. opt (t) Automatically correct pump speed when deviation exceeds ±5%; automatically limit and pop up alarm when MPC output flow exceeds pump hardware limit; automatically cache all current control parameters when the system is powered off, and restore tunneling conditions with one click when powered on.

[0056] Step S5: Status Monitoring and Adaptive Adjustment

[0057] The system classifies faults and configures concurrent scheduling logic: Level 1 faults (main pump failure, main pipeline overpressure) have the highest priority and are handled immediately; Level 2 faults (single-point blockage, local leakage) and Level 3 faults (temperature drift, slight pressure unevenness) are stored in the task queue and executed sequentially after Level 1 faults are eliminated; multiple self-healing actions are not triggered simultaneously to avoid pressure disturbances.

[0058] Clog self-healing grading strategy: When the root mean square (RMS) value of the vibration signal exceeds twice but less than three times the baseline value, it is judged as a sign of mild blockage. The pulse width modulation duty cycle of the solenoid valve at that grease injection point is automatically increased by 5% for 1 second to attempt flushing and clearing. When the RMS value of the vibration signal exceeds three times the baseline value and lasts for more than two seconds, it is judged as a moderate blockage. The backflushing procedure is initiated: the electromagnetic proportional distribution valve corresponding to the grease injection point is closed, the electromagnetic backflushing valve is opened, and the backflushing air source uses 0.4-0.8 MPa compressed air to backflush for 3-10 seconds, repeated 1-3 times. If the vibration value recovers to within twice the baseline value after backflushing, it is judged as a successful recovery, and normal operation continues. If the vibration value is still more than three times the baseline value after the above backflushing procedure is completed, or if it is still ineffective after three executions, it is judged as an unrecoverable blockage. The grease injection point is closed, its distribution coefficient is transferred to the two adjacent grease injection points according to the distance weight, and at the same time, an audible and visual alarm is issued to the control panel and pushed to the remote monitoring terminal.

[0059] Leakage risk boosting: When the pressure P at any point measured by the pressure sensor... i (t) <P soil When (t)+0.2MPa persists for 3 seconds, a leakage risk is identified, and P is adjusted accordingly. target (t) Increase by 20% and increase the allocation coefficient at that point;

[0060] High-temperature viscosity compensation: When the temperature sensor detects that the grease temperature exceeds 50°C, the target pressure is automatically reduced by 0.1 MPa and the grease injection frequency is increased to compensate for the decrease in viscosity.

[0061] Earth pressure change feedforward response: when the earth chamber pressure P soil (t) When the rate of change exceeds 0.1 MPa / s, the K1 gain is increased by 50% to achieve a fast feedforward response.

[0062] Step S6: Emergency Response and Self-Healing

[0063] a. Main pump failure: If the servo motor or variable piston pump fails, the system will automatically switch to the standby pump within 0.2 seconds and display the fault information on the touch screen.

[0064] b. Main pipeline overpressure: If the pressure in the main grease injection pipeline exceeds the set threshold (30-38MPa, default 35MPa), the electrically controlled relief valve will first activate to release pressure. If the pressure continues to rise to the opening pressure of the mechanical safety valve (95% of the upper limit of the threshold), the electromechanical safety valve will automatically open to release pressure to the grease tank.

[0065] c. Grease type switching: When the circumferential pressure unevenness cannot be improved for 30 seconds, the grease type is changed by switching valve to adapt to formation changes;

[0066] d. All fault information and handling results are uploaded to the cloud platform via the remote communication module and pushed to the operator's mobile app for alarm.

[0067] Step S7: Self-learning and parameter evolution

[0068] After each tunneling cycle (1.5 m), the performance indicators for that cycle are calculated: pressure tracking error integral, grease consumption efficiency, and circumferential pressure unevenness. If the indicators for this cycle are better than the historical average, the current control parameters are stored in the experience base in memory. Every morning, using successful samples from the experience base, the adaptive gain coefficients K1, K2, K3 and the model predictive control weight matrix are updated through an incremental learning algorithm. Simultaneously, the federated learning global model is downloaded from the cloud via a remote communication module to achieve cross-site knowledge transfer. The cloud-based federated learning is equipped with security mechanisms: locally uploaded data is anonymized, sensitive information such as project coordinates and shield number is deleted, and only the strata, working conditions, parameters, and performance indicators are retained; the transmission channel is encrypted, and the original construction data does not leave the local edge controller; the cloud uses FedAvg federated average aggregation of model weights and does not summarize the original data; after the global model is downloaded locally, the allocation coefficient benchmark value is automatically corrected based on the local shield diameter and number of sealing channels.

[0069] It should be noted that the present invention also features dual hardware and software safety interlocks, power failure parameter caching and recovery, and multi-fault hierarchical scheduling logic. First-level faults are handled first, while low-level faults are stored in a task queue and self-healing actions are performed sequentially. The dual hardware and software safety interlocks include a two-stage pressure relief circuit consisting of an electrically controlled overflow valve and a mechanical safety valve, and a hardware limiting circuit for the controller's flow output. The power failure parameter caching and recovery is achieved through a non-volatile ferroelectric memory, which archives all process variables within the control cycle at set intervals. Upon power-up, the most recent archived point is automatically loaded and the system is smoothly restored to the operating condition before the power failure.

[0070] Example 1: Fully Automated Grease Injection into High-Water-Pressure Soft Soil Layers

[0071] Project Overview: This is a tunnel project crossing the Yangtze River. The shield diameter is 6.4 m, the water pressure is 0.55 MPa, and the stratum is soft plastic silty clay. The pressure fluctuation range of the soil chamber is 0.45~0.65 MPa. The traditional grout injection system requires manual parameter adjustment every 50 rings of tunneling in this section, and bottom grout leakage still occurs multiple times.

[0072] A control test was added: a traditional fixed PID grease injection system was used as a control group in adjacent sections of the same stratum, and data on pressure fluctuation, grease consumption and leakage frequency were collected and compared simultaneously.

[0073] Implementation process:

[0074] 1) After the device is powered on, it will automatically perform a self-test, which takes about 2 minutes. The touch screen will then display "System ready".

[0075] 2) The tunnel boring machine begins excavation, and the soil chamber pressure transmitter measures P. soil (t) = 0.52 MPa, calculate P target (t) = 1.02 MPa. The pressure sensor measured the circumferential pressure distribution: approximately 0.98 MPa at the top, and only 0.65 MPa at the bottom (6 o'clock), with a non-uniformity ΔP. max =0.33MPa, exceeding the threshold.

[0076] 3) Initiate adaptive decision-making: Demand liposuction rate Q demand The calculated value of (t) is 68 ml / min, and the output Q after model predictive control optimization is... opt (t) = 75 ml / min, partition coefficient β i Distribute 65% of the flow to the bottom three injection points.

[0077] 4) The servo motor drives the variable displacement piston pump, and the electromagnetic proportional distribution valve opens according to the command. After 3 seconds, the bottom pressure rises to 0.98 MPa, the top pressure is 1.05 MPa, and the unevenness drops to 0.07 MPa.

[0078] 5) When tunneling reached the 120th ring, the soil chamber pressure suddenly increased to 0.71 MPa (the strata hardened), with a change rate dP soil / dt=0.19MPa / s, increasing the feedforward gain K1 by 50%, Q opt The transient boost reached 110 ml / min with a pressure response time of 1.2 seconds and no overshoot.

[0079] 6) After this cycle, the system calculates the following performance indicators: the pressure tracking error integral decreases by 8% compared to the previous cycle, and the grease consumption is 0.53 kg / cycle, which is better than the historical average. The control parameters for this cycle are then stored in the experience database.

[0080] Implementation Results: The system achieved continuous tunneling of 800 m with zero manual intervention and no grout leakage incidents. Grease consumption was reduced by 34% compared to traditional equipment. The touchscreen display showed that the system completed 12 automatic mode switches and 2 backwashing self-healing actions (both minor signs of blockage), all of which were successfully resolved. In contrast, the control group's traditional system had a grease consumption of 0.80 kg / ring, a stable circumferential pressure differential of 0.6~0.7 MPa, and experienced 4 shutdowns due to bottom grout leakage.

[0081] Example 2: Self-healing of blockages in gravel and pebble formations and automatic switching between grease and oil

[0082] Project Overview: In a certain subway tunnel section, the stratum is mainly composed of gravelly soil with a particle size of 50~100 mm. The traditional grease injection system experiences blockage of the grease injection pipeline on average once every 100 rings, requiring a shutdown for 2~3 hours to clear the blockage.

[0083] Implementation process:

[0084] 1) When the tunneling reached the 87th ring, the vibration sensor detected that the root mean square value of the vibration of the No. 3 grease injection branch pipe suddenly increased to 4.2 times the baseline value for 2 seconds, which was determined to be a sign of blockage.

[0085] 2) The system automatically initiates the backflushing procedure: The No. 3 electromagnetic proportional distribution valve is closed, and the electromagnetic backflushing valve is opened. The backflushing air source purges in the reverse direction at a pressure of 0.6 MPa for 5 seconds. After the first purge, the vibration value drops to 1.8 times the baseline, and returns to normal after the second purge. The entire process takes 12 seconds and does not affect the continuity of grease injection.

[0086] 3) When tunneling reached the 150th ring, the pressure transmitter in the soil chamber detected increased pressure fluctuations, and the circumferential pressure unevenness persisted for 30 seconds without improvement (the bottom pressure remained 0.15 MPa below the target). This was determined to be due to insufficient grease adaptability caused by changes in formation resistance. The system automatically switched the grease source from the first chamber to the second chamber (higher consistency) via a switching valve. Within 15 seconds after the switch, the circumferential pressure returned to equilibrium.

[0087] 4) The system records this switch as "positive experience" and stores it in memory. During subsequent tunneling, when similar pressure characteristics are detected, the system proactively pre-switch the grease type to avoid pressure fluctuations.

[0088] Implementation Results: After 500 rings of continuous tunneling, the system performed backwashing and self-healing 6 times (100% success rate), automatically switched grease 3 times, and experienced no downtime due to pipeline blockage. The wear of the tail shield brush was reduced by 42% compared to adjacent sections.

[0089] Example 3: Federated Learning for Cross-Site Knowledge Transfer

[0090] Project Overview: The equipment of this invention will be deployed on 5 tunnel boring machines with different geological conditions (soft soil, sand, gravel, composite strata, and rock strata). Each machine will operate independently and will be connected to a cloud-based federated learning server via a remote communication module.

[0091] Implementation process:

[0092] 1) Week 1: Each of the 5 machines used its factory pre-trained parameters, and the average accuracy of the first ring grease injection (the percentage of time with pressure error ≤ ±0.05MPa) was about 72%.

[0093] 2) Week 2: Each machine uploads its successful experiences (working condition characteristics - optimal parameter pairs) from the tunneling process to the cloud after de-identification. The cloud server aggregates and generates a global model.

[0094] 3) Week 3: The global model was distributed to each edge device, and new parameters were loaded. During Week 3, the average grease injection accuracy of the 5 devices jumped to 86%.

[0095] 4) Week 4: Continued independent operation and incremental learning, with an average accuracy of 93%. Among them, after experiencing a large fluctuation in soil chamber pressure, the self-learning module automatically adjusted the K1 gain coefficient of the soft soil stratum equipment, and the pressure overshoot under similar fluctuations subsequently decreased from 0.12 MPa to 0.05 MPa.

[0096] 5) Remote monitoring: Operators can view the pressure cloud map, grease injection trend, and fault warnings of each device in real time via a mobile app. One day, the cloud received a "standby pump switchover" alarm for a device and pushed it to the maintenance personnel. The maintenance personnel arrived on site with the spare part within 2 hours to replace it, avoiding equipment downtime.

[0097] Implementation Results: Federated learning enabled multiple devices to evolve together, improving the initial grease injection accuracy at new construction sites by approximately 29%. Remote monitoring reduced the average fault response time from hours to within 5 minutes.

[0098] Example 4: Adaptive Compensation for Extreme Temperatures

[0099] Project Overview: A tunnel under construction in northern China during winter, with an ambient temperature of -15℃ and an initial grease temperature of -10℃. During grease injection, the grease gradually warms to 40℃. Traditional systems suffer from viscosity variations leading to grease volume deviations exceeding ±20%.

[0100] Implementation process:

[0101] 1) When the equipment starts up, the temperature sensor measures the grease temperature to be -10℃. The temperature-viscosity compensation model is then invoked, and P... target It automatically increases the pressure by 0.15 MPa to compensate for the flow resistance caused by high viscosity.

[0102] 2) During tunneling, the heating function of the dual-chamber constant-temperature oil tank is activated to maintain the grease temperature at 40±2℃. A temperature sensor monitors the temperature in real time, and as the temperature rises from -10℃ to 40℃, the K3 gain coefficient is dynamically adjusted to ensure the required grease injection rate Q is achieved. demand Smooth change.

[0103] 3) The measured flow deviation was always less than ±3%, and the pressure fluctuation in the sealed cavity was less than ±0.05 MPa, while the flow deviation of the bypass control system was as high as ±18%, and the pressure fluctuation was ±0.2 MPa.

[0104] Implementation results: The temperature adaptive function ensures that the grease injection accuracy meets the construction requirements under extreme temperature difference conditions, avoiding the risk of seal failure caused by viscosity changes.

[0105] Example 5: Verification of Adaptation to Large-Diameter Composite Formations

[0106] The tunnel boring machine (TBM) has a diameter of 12.2m and is located in a composite geological formation. After the system inputs the machine parameters, it automatically increases the flow rate distributed at the bottom 6 foundation points by 18%, ΔP. safe The pressure was increased to 0.7MPa; the circumferential pressure difference remained stable at ≤0.09MPa throughout the tunneling process, and the grease consumption per unit area was only 41% higher than that of an 8m shield tunnel in the same stratum, which is lower than the 60% increase of traditional equipment, verifying the effectiveness of the adaptive correction logic of the machine model.

[0107] This invention utilizes circumferential multi-sensor fusion sensing and digital twin reconstruction to dynamically optimize grease injection parameters based on model predictive control (MPC), achieving pressure overshoot ≤6% and response time ≤2 seconds. Circumferential intelligent distribution reduces the pressure difference between the bottom and top from 0.7MPa to within 0.08MPa, eliminating localized grout leakage. It possesses self-healing capabilities for pipeline blockages (backflushing success rate >80%), automatic switching between dual greases, and redundancy between main and backup pumps, reducing unplanned downtime by 90%. Built-in incremental learning and cloud-based federated learning mechanisms make the equipment increasingly intelligent with use, increasing initial accuracy at new construction sites from 72% to 93%. Grease consumption is reduced by 30%~38%, and construction costs per kilometer are reduced by over 500,000 yuan. Its modular design allows for adaptation to existing tunnel boring machines, resulting in a wide range of applications and significant overall benefits.

[0108] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for adaptive grease injection for shield tail sealing, characterized in that, include: Step 1: Synchronously collect status sensing parameters of the construction area at a set frequency; The state sensing parameters include: construction area pressure, grease temperature, soil chamber pressure, mass change rate, instantaneous flow rate, and vibration signal; Step 2: Calculate the target pressure, circumferential pressure non-uniformity, and required grease injection rate based on the status sensing parameters; Step 3: Based on the target pressure, circumferential pressure unevenness, and required grease injection rate, calculate the optimal grease injection rate and the distribution coefficient of each grease injection point; Step 4: Control the servo motor and variable piston pump of the shield tail sealing system to operate at the optimal grease injection rate, and adjust the opening of each electromagnetic proportional distribution valve according to the distribution coefficient.

2. The adaptive grease injection method for shield tail sealing grease according to claim 1, characterized in that, In step 2, the formula for calculating the required grease injection rate is: Where K1, K2, and K3 are adaptive gain coefficients; For the required grease injection rate; P target For target pressure; P soil For earth chamber pressure; For time; Circumferential pressure; Q0 is the baseline grease injection rate; µ(T) is the temperature-viscosity compensation function. α is the viscosity temperature coefficient, T0 is the reference temperature; P target =P soil +ΔP safe ΔP safe For safety margin.

3. The adaptive grease injection method for shield tail sealing grease according to claim 2, characterized in that, In step 3, the model predictive control algorithm is used to predict the pressure evolution over the next 5 seconds and solve for the optimal grease injection rate sequence that minimizes the weighted sum of the pressure tracking error and grease consumption. The control period is 100 ms.

4. The adaptive grease injection method for shield tail sealing grease according to claim 3, characterized in that, In step 3, the model predictive control algorithm is established based on the circumferential pressure sequence, soil chamber pressure sequence, grease injection rate sequence, and grease temperature sequence in the historical tunneling data of the tunnel boring machine, through a system identification method.

5. A shield tail sealing grease adaptive grease injection system, characterized in that, include: The parameter acquisition module is used to synchronously acquire status sensing parameters of the construction area at a set frequency. The state sensing parameters include: construction area pressure, grease temperature, soil chamber pressure, mass change rate, instantaneous flow rate, and vibration signal; The parameter calculation module is used to calculate the target pressure, circumferential pressure unevenness, and required grease injection rate based on the state-sensing parameters. The working parameter calculation module is used to solve the optimal grease injection rate and the distribution coefficient of each grease injection point based on the target pressure, circumferential pressure unevenness and required grease injection rate. The operation control module is used to control the servo motor and variable piston pump of the shield tail sealing system to operate at the optimal grease injection rate, and to adjust the opening of each electromagnetic proportional distribution valve according to the distribution coefficient.

6. The adaptive grease injection system for shield tail sealing as described in claim 5, characterized in that, In the parameter calculation module, the formula for calculating the required grease injection rate is: Where K1, K2, and K3 are adaptive gain coefficients; For the required grease injection rate; P target For target pressure; P soil For earth chamber pressure; For time; Circumferential pressure; Q0 is the baseline grease injection rate; µ(T) is the temperature-viscosity compensation function. α is the viscosity temperature coefficient, T0 is the reference temperature; P target =P soil +ΔP safe ΔP safe For safety margin.

7. The adaptive grease injection system for shield tail sealing as described in claim 6, characterized in that, In the working parameter calculation module, the model predictive control algorithm is used to predict the pressure evolution in the next 5 seconds and solve for the optimal grease injection rate sequence that minimizes the weighted sum of the pressure tracking error squared and grease consumption. The control period is 100 ms.

8. The adaptive grease injection system for shield tail sealing as described in claim 7, characterized in that, In the working parameter calculation module, the prediction model in the model prediction control algorithm is established based on the circumferential pressure sequence, soil chamber pressure sequence, grease injection rate sequence, and grease temperature sequence in the historical tunneling data of the tunnel boring machine, through the system identification method.

9. An electronic device comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, characterized in that, When the computer program is executed by the processor, it implements the steps in the adaptive grease injection method for shield tail sealing grease as described in any one of claims 1-4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps in the adaptive grease injection construction method for shield tail sealing grease as described in any one of claims 1-4.

Images