Construction method for inclined support of foundation pit

Abstract

The invention discloses a foundation pit inclined support construction method, and belongs to the technical field of foundation pit support engineering, the foundation pit inclined support construction method comprises the following steps: S1, based on a depth layering initial inclination angle parameter frame, combining with soil layer wave velocity dynamic correction support geometric morphology and axial force reference; s2, checking a three-dimensional laser grid and an ink line, outputting a tolerance drawing, and carrying out over-limit locking on the step S3; s3, excavation is conducted according to the sine ratio, excavation is stopped due to seepage abnormity, and the inclined strut installation window period is dynamically compressed according to axial force data; s4, performing staged torque control; s5, the axial force is monitored through wireless sensing, and + / -15% deviation linkage is carried out; S6, grouting compensation is carried out, and a detachable sleeve reuse monitoring unit is adopted; s6, linearly associating the grouting pressure with the depth, performing 0.2 MPa / stage pressurization when the foundation displacement exceeds the limit, and interlocking a redundancy threshold value in S7; s7, fusing multi-source data to construct a three-dimensional acceptance model, and automatically generating an evaluation report by comparing with a standard; the method has the beneficial effects that millimeter-level precision construction control is achieved, and the effect of whole-cycle cost optimization is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of foundation pit support engineering, and more specifically to a foundation pit oblique support construction method. Background Art

[0002] As the development of urban underground space develops towards the direction of deep, large and complex, the traditional foundation pit oblique support technology faces severe challenges. The existing methods mostly adopt fixed inclination design, ignoring the impact of dynamic changes in geological conditions on the support system, resulting in problems such as stress concentration or insufficient stiffness in the support structure. In conventional construction, the setting of pre-axial force parameters relies on empirical formulas, lacks dynamic adaptation to soil characteristics, and often causes support deformation to exceed the standard in soft soil or high water level formations. The positioning system mostly adopts a single manual line setting method, and the cumulative error is difficult to control, especially in deep foundation pits, which is prone to millimeter-level deviation accumulation effects. The existing monitoring system is mostly operated in independent modules, with serious data island phenomenon, and it is impossible to achieve real-time closed-loop optimization of construction parameters. More importantly, the traditional method lacks dynamic coupling control of the support installation window period and excavation progress, resulting in an increased risk of support lag. According to statistics, about 23% of foundation pit accidents are caused by the response lag of the support system. These defects seriously restrict the safety and economy of deep foundation pit construction, and it is urgent to establish a dynamic control system with intelligent decision-making capabilities. Summary of the invention

[0003] The present invention proposes a method for the construction of oblique support for foundation pits, which aims to achieve adaptive adjustment of construction parameters with geological conditions and project progress by establishing a closed-loop system of dynamic compensation for inclination angles, multi-source data interlocking control and intelligent decision-making optimization, thereby significantly improving the safety and construction efficiency of deep foundation pit support structures.

[0004] Technical solution: A foundation pit oblique support construction method, comprising the following steps: S1. A dynamic adjustment mechanism for the inclination angle is established based on the design depth of the foundation pit. The initial reference inclination angle is set to 35 degrees, and a gradient compensation of 0.2 degrees is performed for every 1 meter increase in depth. The reference inclination angle refers to the inclination angle value of the foundation pit oblique support construction. The reference inclination angle is based on the reference value of the oblique support installation angle after dynamic adjustment based on the design depth of the foundation pit and geological conditions. The pre-axial force reference value is calculated synchronously according to the depth logarithmic compensation law, in which the thickness of the embedded steel plate and the bearing characteristics of the high-strength bolts are mechanically adapted and calculated; when the shear wave velocity detection value of the soil layer is lower than the critical threshold, the geological compensation mechanism is triggered to make a 5-degree additional correction to the reference inclination angle, forming a dual reference system including geometric parameters and mechanical parameters; S2. Generate a 3D laser positioning grid based on the geometric reference parameters output by S1. After the 3D laser positioning grid is verified with the manual ink line reference line, a digital construction drawing with a tolerance threshold is output, and the deviation data is transmitted to the S4 weld angle compensation module and the S7 acceptance model in real time. When the limit is exceeded, an early warning is triggered and the subsequent S3 excavation process is locked; S3. According to the reference inclination angle determined in step S1 and corrected by depth gradient compensation and geological conditions, the single-layer excavation depth is calculated according to the positive proportional relationship of its sine value. When the exposed surface film monitoring system detects abnormal seepage or stress mutation, the hydraulic excavation device is automatically blocked; at the same time, according to the S1 reference inclination angle parameter and the S5 real-time axial force monitoring data, the duration of the diagonal brace installation window period is dynamically adjusted through a time-varying algorithm, and the window period is reduced exponentially as the excavation depth increases to ensure the minimum safety duration; S4. Build a dynamic adjustment model and perform staged torque control: Based on the S2 positioning deviation data stream (lateral deviation Δx, longitudinal deviation Δy) and the S3 structural strain rate ε, a dynamic calculation model for the weld angle compensation is established: in: is the horizontal and vertical characteristic length of the foundation pit, is the reference strain rate (set to 0.5 μstrain / s), is the positioning deviation weight coefficient (dynamic range 0.5~1.5), is the strain rate weight coefficient (dynamic range 0.8~1.2); the model fits the parameters in real time through the recursive least squares method and outputs the weld angle compensation Q (in millimeters); performs staged torque control: initial tightening stage: apply torque Secondary preload stage: adjust the torque according to the compensation amount Q in The maximum compensation allowed (set to 10mm); Final tightening stage: applying torque If the residual stress difference exceeds 12%, the residual stress distribution is tested by ultrasonic wave. Reverse correction , the update formula is: in, represents the old value of the weight coefficient, Represents the new value of the weight coefficient; S5. Receive the S1 pre-loaded axial force design value to build a stress monitoring network, collect the diagonal brace axial force data in a grid-like manner through the wireless sensor unit, trigger the compensation mechanism when the real-time axial force deviates from the design value by ±15%, and the compensation amount is based on the S6 grouting pressure gradient to establish a nonlinear mapping relationship; when the support system is dismantled, the stress monitoring unit is detachable and reused through the protective sleeve, and the reuse status data is transmitted back to the S7 acceptance evaluation system in real time; S6. The initial grouting pressure establishes a linear relationship based on the S1 depth parameter. When the S5 foundation displacement thermal distribution map exceeds the limit signal, the graded boost grouting mode is started, and the pressure gradient increases by 0.2MPa / level. The upper limit of the pressure establishes a dynamic interlocking mechanism with the displacement redundancy threshold of the S7 acceptance model; S7. Integrate S2 positioning topology data and S5 axial force cloud map to build a three-dimensional acceptance model, spatially align S2 positioning topology data with S5 axial force cloud map, and automatically identify non-compliant areas by comparing with design standards; For non-compliant areas, give priority to using S1 benchmark inclination correction strategy to add oblique support, and the second option is to trigger S6 grouting reinforcement mechanism; Iteratively optimize the S1 initial parameter group through machine learning algorithm to form a closed-loop control system with adaptive parameters.

[0005] Preferably, the S1 specifically comprises the following steps: S1-1. Depth logarithmic compensation: The natural logarithm of the foundation pit depth is used as the calculation base. When the depth reaches the 5-meter reference threshold, the pre-loaded axial force value is compensated at a gradient of 15% for every 1 unit increase in the natural logarithm. The maximum compensation value shall not exceed 200% of the design value, and the output parameter shall be used as the design basis for the pre-loaded axial force in step S5. S1-2. Geological compensation triggering. In the soil shear wave velocity detection, when the wave velocity values ​​of three consecutive detection points are lower than 80% of the critical threshold, the benchmark inclination compensation mechanism is triggered. The compensation amplitude is 5 degrees and the cumulative number of times does not exceed twice. The compensation result is synchronized in real time to the inclination control module of step S3.

[0006] Preferably, the S2 specifically includes the following steps: S2-1. 3D grid construction, with the center of the bottom surface of the foundation pit as the coordinate origin, laser positioning points are arranged along the reference inclination direction according to the sinusoidal distribution density, the distance between adjacent points is inversely proportional to the square root of the foundation pit depth, and the distance is reduced by 20% for every 5 meters increase in depth, and the generated data is used as the reference input for weld angle compensation in step S4; S2-2. Dynamic tolerance adjustment: when the foundation pit depth exceeds 15 meters, the lateral deviation threshold is dynamically adjusted at a ratio of 0.3 mm per meter of depth, and the longitudinal deviation threshold is maintained at 120% of the initial value. Exceeding limit data directly triggers the acceptance model alarm mechanism of step S7.

[0007] Preferably, the S3 specifically includes the following steps: S3-1. Window period calculation, establish an exponential decay relationship between excavation depth and installation window period, the initial 4-hour window period is shortened to 60% of the original length when the depth reaches 8 meters, the attenuation coefficient takes the natural logarithm of the depth value, and the output parameter controls the axial force collection frequency of step S5; S3-2. Safety time guarantee: by superimposing the minimum response time of the hydraulic system of 0.8 hours and the structural strain buffer time of 0.4 hours, the lower limit of the window period is determined to be 1.2 hours and cannot be exceeded. The limit data is transmitted back to the positioning verification system of step S2 in real time.

[0008] Preferably, the S4 specifically comprises the following steps: S4-1. Multi-level torque control: Initial tightening stage: Apply 50% of the design torque ( ), maintain stress relaxation for 10 minutes, secondary preload stage: based on the weld angle compensation Q output by the dynamic adjustment model, according to the formula: Adjust the torque, where Q is calculated in real time by the S4 main step model; if the strain rate ε exceeds the reference value , then the additional strain compensation term is superimposed, and the torque upper limit is 150% of the design value. Final tightening stage: Apply 100% design torque ( ); S4-2. Residual stress verification: After final tightening, ultrasonic testing of residual stress distribution is used, requiring the difference between any two points to be ≤12%; When the limit is exceeded, the S5 axial force monitoring unit is called for cross-validation and the model error is calculated: By formula Reverse correction weight coefficient , the updated parameters are written into the dynamic adjustment model; The verification results are synchronized to the S7 acceptance model for subsequent construction section parameter optimization.

[0009] Preferably, the S5 specifically includes the following steps: S5-1. Compensation trigger: when the real-time axial force value deviates from the design value by plus or minus 15%, the compensation mechanism linked to step S6 is activated. The compensation amount is calculated according to the square root ratio of the grouting pressure gradient, and the trigger threshold is dynamically adjusted according to the S1 pre-axial force reference value; S5-2. Unit reuse. When the support is removed, the monitoring unit is double-index calibrated, requiring the calibration error to be less than 0.8% and the protective structure to be intact. The reuse status data is encrypted and transmitted to the S7 acceptance system and fed back to the positioning accuracy optimization algorithm in step S2.

[0010] Preferably, the S6 specifically comprises the following steps: S6-1. Pressure grading control: when the foundation displacement reaches the yellow warning zone of the thermal distribution map, 0.2MPa foundation grouting is started. After entering the red warning zone, the pressure is increased by 0.2MPa per level. The upper limit of pressure is 125% of the S7 displacement redundancy threshold. The warning grading standard is inherited from the S2 positioning deviation historical data; S6-2. Pressure interlock. When the grouting pressure gradient is increased to the third level, the S5 axial force data is automatically called to establish a pressure-axial force feedback loop. The pressure increase is adjusted to 80% of the previous value, and the feedback loop parameters are synchronized to the S3 excavation control timing table.

[0011] Preferably, the S7 specifically includes the following steps: S7-1. 3D model construction and deviation assessment: Based on the coordinate data of the S2 laser positioning grid (accuracy ±0.3mm) and the S5 axial force monitoring data (sampling frequency ≥1Hz), the following method is used to construct the 3D acceptance model: S7-1-1. The foundation pit support structure is meshed using the Delaunay triangulation algorithm. The mesh density varies with depth. ( , Dynamic encryption for depth; S7-1-2. Use Kriging interpolation method to generate continuous axial force cloud map, the calculation formula is: in, is the weight coefficient calculated for the semivariogram, is the drift term in the depth direction; Representing coordinates The predicted axial force value at ; Indicates the lateral coordinate of the foundation pit; Indicates the longitudinal coordinate of the foundation pit; z indicates the vertical depth of the foundation pit Indicates The measured axial force values ​​of the monitoring points are obtained through gridding collection of the wireless sensor units in step S5; S7-1-3. Establish a deviation assessment matrix: when It is judged as a non-compliant area; in: It indicates the lateral positioning deviation of the foundation pit support structure, that is, the difference between the actual position and the design coordinates in the lateral direction; Indicates longitudinal positioning deviation; represents the lateral deviation threshold; represents the longitudinal deviation threshold; Indicates the measured axial force value; Indicates the design axial force value; S7-2. Reinforcement decision driven by curvature analysis, cubic spline curvature analysis of non-compliant areas: S7-2-1. Fitting axial force distribution curve And calculate the curvature: in, Represents the curvature of the axial force distribution curve; : Axial force distribution function The first derivative of ; Represents the axial force distribution function The second derivative of S7-2-2. When When S1 reference inclination correction is triggered first in, represents the corrected reference inclination; Indicates the reference inclination before correction; : Sign function, the direction of inclination adjustment is determined according to the sign of the difference between the measured axial force and the designed axial force: If , , the inclination angle increases by 5°; if , , the inclination angle is reduced by 5°.

[0012] S7-2-3. When When S6 secondary grouting is triggered, the grouting pressure is in, Indicates grouting pressure; Make dynamic adjustments; S7-3. Parameter adaptive optimization and verification: S7-3-1. Use the exponential moving average algorithm (EMA) to update the initial parameters of S1. The formula is: in, is the updated initial reference inclination, is the initial reference inclination before updating, is the actual reference inclination angle used in the current construction section, is the attenuation factor, is the weight of the current data, n represents the nth construction; S7-3-2. The updated parameters are simulated and verified by the S2 positioning system to ensure that the three-dimensional laser positioning deviation is ≤0.5mm / m (horizontally) and ≤1.0mm / m (vertically). After verification, they are synchronized to the construction control module; S7-3-3. Use support vector regression (SVR) to train historical data and optimize the weights of the deviation evaluation matrix, with positioning deviation accounting for 60% and axial force deviation accounting for 40%, forming a parameter optimization closed loop.

[0013] Compared with the prior art, the advantages of the present invention are: (1) Dynamic inclination compensation mechanism: By integrating geological parameters and engineering depth characteristics, a gradient adaptive adjustment model is constructed. The support inclination can be dynamically optimized according to the mechanical properties of the stratum and the excavation depth, effectively balancing the stress distribution under different working conditions, significantly improving the adaptability of the support structure to complex strata, and solving the problem of local stress concentration caused by traditional fixed inclination design.

[0014] (2) Closed-loop control system with multi-source data interlocking: Integrate multi-dimensional data streams such as high-precision positioning, real-time stress monitoring, and grouting pressure feedback to establish a dynamic joint adjustment mechanism for construction parameters, realize coordinated control of the entire process from excavation to support, break through the technical limitations of parameter adjustment lag and data silos in traditional methods, and ensure real-time matching of construction response and geological changes.

[0015] (3) Intensive design of reusable monitoring units and intelligent optimization of construction sequence control: The modular sensor structure and protective sleeve integration technology are used to realize the rapid disassembly and reuse of the monitoring unit, which greatly reduces the equipment loss cost. At the same time, through encrypted data transmission and long-term performance tracking, continuous data support is provided for construction parameter optimization. Based on the dynamic relationship model between excavation depth and structural safety threshold, the support installation window period is automatically adjusted. Combined with the dual guarantee mechanism of hydraulic system response and structural buffer time, the risk of support delay is effectively avoided and the precise synchronization of the support system and excavation progress is ensured. DETAILED DESCRIPTION

[0016] Embodiment, a foundation pit oblique support construction method, comprising the following steps: S1. A dynamic adjustment mechanism for the inclination angle is established based on the design depth of the foundation pit. The initial reference inclination angle is set to 35 degrees, and a gradient compensation of 0.2 degrees is performed for every 1 meter increase in depth. The reference inclination angle refers to the reference value of the inclined support installation angle after dynamic adjustment based on the design depth of the foundation pit and geological conditions. The pre-axial force reference value is calculated based on the depth logarithmic compensation law, in which the thickness of the embedded steel plate and the bearing characteristics of the high-strength bolts are mechanically adapted and calculated. When the shear wave velocity detection value of the soil layer is lower than the critical threshold, the geological compensation mechanism is triggered to make a 5-degree additional correction to the reference inclination angle, forming a dual reference system including geometric parameters and mechanical parameters. S2. Generate a 3D laser positioning grid based on the geometric reference parameters output by S1. After the 3D laser positioning grid is verified with the manual ink line reference line, a digital construction drawing with a tolerance threshold is output, and the deviation data is transmitted to the S4 weld angle compensation module and the S7 acceptance model in real time. When the limit is exceeded, an early warning is triggered and the subsequent S3 excavation process is locked; S3. According to the reference inclination angle determined in step S1 and corrected by depth gradient compensation and geological conditions, the single-layer excavation depth is calculated according to the positive proportional relationship of its sine value. When the exposed surface film monitoring system detects abnormal seepage or stress mutation, the hydraulic excavation device is automatically blocked; at the same time, according to the S1 reference inclination angle parameter and the S5 real-time axial force monitoring data, the duration of the diagonal brace installation window period is dynamically adjusted through a time-varying algorithm, and the window period is reduced exponentially as the excavation depth increases to ensure the minimum safety duration; S4. Build a dynamic adjustment model and perform staged torque control: Based on the S2 positioning deviation data stream (lateral deviation Δx, longitudinal deviation Δy) and the S3 structural strain rate ε, a dynamic calculation model for the weld angle compensation is established: in: is the horizontal and vertical characteristic length of the foundation pit, is the reference strain rate (set to 0.5 μstrain / s), is the positioning deviation weight coefficient (dynamic range 0.5~1.5), is the strain rate weight coefficient (dynamic range 0.8~1.2); the model fits the parameters in real time through the recursive least squares method and outputs the weld angle compensation Q (in millimeters); performs staged torque control: initial tightening stage: apply torque Secondary preload stage: adjust the torque according to the compensation amount Q in The maximum compensation allowed (set to 10mm); Final tightening stage: applying torque If the residual stress difference exceeds 12%, the residual stress distribution is tested by ultrasonic wave. Reverse correction , the update formula is: in, represents the old value of the weight coefficient, Represents the new value of the weight coefficient; S5. Receive the S1 pre-loaded axial force design value to build a stress monitoring network, collect the diagonal brace axial force data in a grid-like manner through the wireless sensor unit, trigger the compensation mechanism when the real-time axial force deviates from the design value by ±15%, and the compensation amount is based on the S6 grouting pressure gradient to establish a nonlinear mapping relationship; when the support system is dismantled, the stress monitoring unit is detachable and reused through the protective sleeve, and the reuse status data is transmitted back to the S7 acceptance evaluation system in real time; S6. The initial grouting pressure establishes a linear relationship based on the S1 depth parameter. When the S5 foundation displacement thermal distribution map exceeds the limit signal, the graded boost grouting mode is started, and the pressure gradient increases by 0.2MPa / level. The upper limit of the pressure establishes a dynamic interlocking mechanism with the displacement redundancy threshold of the S7 acceptance model; S7. Integrate S2 positioning topology data and S5 axial force cloud map to build a three-dimensional acceptance model, spatially align S2 positioning topology data with S5 axial force cloud map, and automatically identify non-compliant areas by comparing with design standards; For non-compliant areas, give priority to using S1 benchmark inclination correction strategy to add oblique support, and the second option is to trigger S6 grouting reinforcement mechanism; Iteratively optimize the S1 initial parameter group through machine learning algorithm to form a closed-loop control system with adaptive parameters.

[0017] The S1 specifically includes the following steps: S1-1. Depth logarithmic compensation: The core is to establish a nonlinear compensation relationship between the foundation pit depth and the pre-loaded axial force. The natural logarithmic value of the foundation pit depth is used as the calculation base. When the foundation pit depth reaches the benchmark threshold of 5 meters, the system starts the logarithmic compensation mechanism. For example, when the foundation pit depth increases from 5 meters to 6 meters (the natural logarithmic increment is approximately 0.182), the pre-loaded axial force value needs to be increased by a gradient of 15% per unit logarithmic increment. In specific implementation, if the design pre-loaded axial force benchmark is 1000kN, the axial force increases to 1150kN for every increase of 1 natural logarithmic unit in depth (such as from 5 meters to about 8.2 meters), but the total compensation does not exceed 200% of the design value (ie 2000kN). This mechanism effectively responds to the characteristics of the nonlinear increase in lateral pressure of deep soil, while avoiding the waste of resources caused by over-compensation; S1-2. Geological compensation is triggered, and the shear wave velocity of the soil layer is a key indicator reflecting the geological conditions. When the wave velocity values ​​of three consecutive detection points are lower than 80% of the critical threshold, it indicates that there are significantly weak or loose areas in the soil layer. At this time, the system automatically triggers the inclination compensation mechanism and increases the reference inclination by 5 degrees. For example, if the initial inclination is 35 degrees, it is adjusted to 40 degrees after the first compensation, and the upper limit of the cumulative number is two times, that is, the maximum compensation amplitude is 10 degrees. The corrected inclination parameters are synchronized to the inclination control module of S3 in real time to ensure that the installation angle of the diagonal brace is dynamically adapted to the geological conditions to prevent support failure due to insufficient soil strength.

[0018] The S2 specifically includes the following steps: S2-1. Three-dimensional grid construction. The construction of the three-dimensional laser positioning grid takes the center of the bottom surface of the foundation pit as the coordinate origin, and distributes the laser positioning points according to the density of the sine curve along the reference inclination angle. The distance between adjacent points is inversely proportional to the square root of the depth of the foundation pit. For example, when the depth is 10 meters, the spacing is set to 1 / √10 (about 31.6%) of the initial value. For every 5 meters increase in depth, the spacing is further reduced by 20%. This distribution method can optimize the density of positioning points and ensure the measurement accuracy of deep areas. The generated three-dimensional grid data is used as the reference input for S4 weld angle compensation, providing a spatial reference for subsequent deviation correction; S2-2. Dynamic tolerance adjustment: When the depth of the foundation pit exceeds 15 meters, the lateral deviation threshold is dynamically adjusted by reducing it by 0.3 mm per meter of depth. For example, at a depth of 16 meters, the lateral tolerance is reduced by 0.3 mm, while the longitudinal tolerance remains at 120% of the initial value. If an over-limit deviation is detected (such as a lateral deviation exceeding the adjusted threshold), the system directly triggers the alarm mechanism of the S7 acceptance model and suspends the S3 excavation process until manual review is completed. This mechanism balances the high requirements for lateral stability and the controllability of longitudinal deformation in deep foundation pit construction.

[0019] The S3 specifically includes the following steps: S3-1. Window period calculation, the installation window period is exponentially decaying with the excavation depth, and the initial window period is 4 hours. When the depth reaches 8 meters, the window period is shortened to 60% of the original length (i.e. 2.4 hours), and the attenuation coefficient takes the natural logarithm of the depth value. For example, at a depth of 12 meters, the attenuation coefficient is the natural logarithm 12≈2.485, and the window period is further reduced to 1.2 hours. This algorithm ensures the efficiency of deep construction, and at the same time ensures the timeliness of data by dynamically adjusting the S5 axial force collection frequency (such as from once per hour to once every 30 minutes); S3-2. Safety time guarantee, the lower limit of the window period is determined by the superposition of the minimum response time of the hydraulic system (0.8 hours) and the structural strain buffer time (0.4 hours), that is, 1.2 hours. For example, when the depth causes the calculated window period value to be less than 1.2 hours, the system is forced to lock it to 1.2 hours and feedback the limit value data to the S2 positioning verification system to prevent equipment response lag or structural instability due to insufficient time.

[0020] The S4 specifically comprises the following steps: S4-1. Multi-level torque control: Initial tightening stage: Apply 50% of the design torque ( ), maintain stress relaxation for 10 minutes, secondary preload stage: based on the weld angle compensation Q output by the dynamic adjustment model, according to the formula: Adjust the torque, where Q is calculated in real time by the S4 main step model; if the strain rate ε exceeds the reference value , then the additional strain compensation term is superimposed, and the torque upper limit is 150% of the design value. Final tightening stage: Apply 100% design torque ( ); S4-2. Residual stress verification: After the final tightening is completed, ultrasonic testing of residual stress distribution is used, and the difference between any two points is required to be ≤12%; When the limit is exceeded, the S5 axial force monitoring unit is called for cross-validation and the model error is calculated: By formula Reverse correction weight coefficient , the updated parameters are written into the dynamic adjustment model; The verification results are synchronized to the S7 acceptance model for subsequent construction section parameter optimization.

[0021] The S5 specifically includes the following steps: S5-1. Compensation trigger: when the real-time axial force deviates from the design value by ±15% (e.g. the design axial force is 1000kN, and the measured value reaches 1150kN or 850kN), the system activates the compensation mechanism linked to S6. The compensation amount is calculated according to the square root ratio of the grouting pressure gradient. For example, if the grouting pressure gradient is 0.2MPa / level, the compensation amount is a multiple of the square root of the pressure gradient (e.g. √0.2≈0.447 times), ensuring the physical correlation between the grouting amount and the axial force deviation; S5-2. Unit reuse, dual-index calibration of monitoring units when supports are removed: calibration error must be less than 0.8%, and protective structure must be intact. For example, if the calibration error of a unit is 0.5% and the protective sleeve is intact, it is marked as reusable, and its status data is encrypted and transmitted to the S7 acceptance system, and fed back to the S2 positioning accuracy optimization algorithm to improve the measurement accuracy of subsequent construction.

[0022] The S6 specifically comprises the following steps: S6-1. Pressure grading control, grouting pressure is graded and adjusted according to the thermal distribution map of foundation displacement. When the foundation displacement enters the yellow warning zone (such as the displacement reaches 20mm), 0.2MPa foundation grouting is started; after entering the red warning zone (displacement reaches 30mm), the pressure increases by 0.2MPa per level, and the upper limit of the pressure is 125% of the S7 displacement redundancy threshold (if the threshold is 1.6MPa, the upper limit is 2.0MPa). The warning grading standard is inherited from the S2 historical deviation data to ensure that the grouting strategy matches the geological risk level; S6-2. Pressure interlock: When the grouting pressure is raised to the third level (0.6MPa), the system calls the S5 axial force data to establish a pressure-axial force feedback loop, and the pressure increase is adjusted to 80% of the previous value (for example, if the previous increment is 0.2MPa, the subsequent increment is adjusted to 0.16MPa) to avoid soil splitting caused by a sudden increase in pressure. The feedback loop parameters are synchronized to the S3 excavation control schedule to optimize the collaborative efficiency of deep excavation and grouting.

[0023] The S7 specifically comprises the following steps: S7-1. 3D model construction and deviation assessment: Based on the coordinate data of the S2 laser positioning grid (accuracy ±0.3mm) and the S5 axial force monitoring data (sampling frequency ≥1Hz), the following method is used to construct the 3D acceptance model: S7-1-1. The foundation pit support structure is meshed using the Delaunay triangulation algorithm. The mesh density varies with depth. ( , Dynamic encryption for depth; S7-1-2. Use Kriging interpolation method to generate continuous axial force cloud map, the calculation formula is: in, is the weight coefficient calculated for the semivariogram, is the drift term in the depth direction; Representing coordinates The predicted axial force value at ; Indicates the lateral coordinate of the foundation pit; Indicates the longitudinal coordinate of the foundation pit; z indicates the vertical depth of the foundation pit Indicates The measured axial force values ​​of the monitoring points are obtained through gridding collection of the wireless sensor units in step S5; S7-1-3. Establish a deviation assessment matrix: when It is judged as a non-compliant area; in: It indicates the lateral positioning deviation of the foundation pit support structure, that is, the difference between the actual position and the design coordinates in the lateral direction; Indicates longitudinal positioning deviation; represents the lateral deviation threshold; represents the longitudinal deviation threshold; Indicates the measured axial force value; Indicates the design axial force value; S7-2. Reinforcement decision driven by curvature analysis, cubic spline curvature analysis of non-compliant areas: S7-2-1. Fitting axial force distribution curve And calculate the curvature: in, Represents the curvature of the axial force distribution curve; : Axial force distribution function The first derivative of ; Represents the axial force distribution function The second derivative of S7-2-2. When When S1 reference inclination correction is triggered first in, represents the corrected reference inclination; Indicates the reference inclination before correction; : Sign function, the direction of inclination adjustment is determined according to the sign of the difference between the measured axial force and the designed axial force: If , , the inclination angle increases by 5°; if , , the inclination angle is reduced by 5°.

[0024] S7-2-3. When When S6 secondary grouting is triggered, the grouting pressure is Make dynamic adjustments; S7-3. Parameter adaptive optimization and verification: S7-3-1. Use the exponential moving average algorithm (EMA) to update the initial parameters of S1. The formula is: in, is the updated initial reference inclination, is the initial reference inclination before updating, is the actual reference inclination angle used in the current construction section, is the attenuation factor, is the weight of the current data, n represents the nth construction; S7-3-2. The updated parameters are simulated and verified by the S2 positioning system to ensure that the three-dimensional laser positioning deviation is ≤0.5mm / m (horizontally) and ≤1.0mm / m (vertically). After verification, they are synchronized to the construction control module; S7-3-3. Use support vector regression (SVR) to train historical data and optimize the weights of the deviation evaluation matrix, with positioning deviation accounting for 60% and axial force deviation accounting for 40%, forming a parameter optimization closed loop.

[0025] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A foundation pit oblique support construction method, characterized in that: The following steps are involved: S1. A dynamic adjustment mechanism for the inclination angle is established based on the design depth of the foundation pit. The initial reference inclination angle is set to 35 degrees, and a gradient compensation of 0.2 degrees is performed for every 1 meter increase in depth. The reference inclination angle refers to the inclination angle value of the foundation pit oblique support construction. The reference inclination angle is based on the reference value of the oblique support installation angle after dynamic adjustment based on the design depth of the foundation pit and geological conditions. The pre-axial force reference value is calculated synchronously according to the depth logarithmic compensation law, in which the thickness of the embedded steel plate and the bearing characteristics of the high-strength bolts are mechanically adapted and calculated; when the shear wave velocity detection value of the soil layer is lower than the critical threshold, the geological compensation mechanism is triggered to make a 5-degree additional correction to the reference inclination angle, forming a dual reference system including geometric parameters and mechanical parameters; S2. Generate a 3D laser positioning grid based on the geometric reference parameters output by S1. After the 3D laser positioning grid is verified with the manual ink line reference line, a digital construction drawing with a tolerance threshold is output, and the deviation data is transmitted to the S4 weld angle compensation module and the S7 acceptance model in real time. When the limit is exceeded, an early warning is triggered and the subsequent S3 excavation process is locked; S3. According to the reference inclination angle determined in step S1 and corrected by depth gradient compensation and geological conditions, the single-layer excavation depth is calculated according to the positive proportional relationship of its sine value. When the exposed surface film monitoring system detects abnormal seepage or stress mutation, the hydraulic excavation device is automatically blocked; at the same time, according to the S1 reference inclination angle parameter and the S5 real-time axial force monitoring data, the duration of the diagonal brace installation window period is dynamically adjusted through a time-varying algorithm, and the window period is reduced exponentially as the excavation depth increases to ensure the minimum safety duration; S4. Build a dynamic adjustment model and perform staged torque control: Based on the S2 positioning deviation data stream (lateral deviation Δx, longitudinal deviation Δy) and the S3 structural strain rate ε, a dynamic calculation model for the weld angle compensation is established: , in: is the horizontal and vertical characteristic length of the foundation pit, is the reference strain rate (set to 0.5 μstrain / s), is the positioning deviation weight coefficient (dynamic range 0.5~1.5), is the strain rate weight coefficient (dynamic range 0.8~1.2); the model fits the parameters in real time through the recursive least squares method and outputs the weld angle compensation Q (in millimeters); performs staged torque control: initial tightening stage: apply torque Secondary preload stage: adjust the torque according to the compensation amount Q in The maximum compensation allowed (set to 10mm); Final tightening stage: applying torque If the residual stress difference exceeds 12%, the residual stress distribution is tested by ultrasonic wave. Reverse correction , the update formula is: in, represents the old value of the weight coefficient, Represents the new value of the weight coefficient; S5. Receive the S1 pre-loaded axial force design value to build a stress monitoring network, collect the diagonal brace axial force data in a grid-like manner through the wireless sensor unit, trigger the compensation mechanism when the real-time axial force deviates from the design value by ±15%, and the compensation amount is based on the S6 grouting pressure gradient to establish a nonlinear mapping relationship; when the support system is dismantled, the stress monitoring unit is detachable and reused through the protective sleeve, and the reuse status data is transmitted back to the S7 acceptance evaluation system in real time; S6. The initial grouting pressure establishes a linear relationship based on the S1 depth parameter. When the S5 foundation displacement thermal distribution map exceeds the limit signal, the graded boost grouting mode is started, and the pressure gradient increases by 0.2MPa / level. The upper limit of the pressure establishes a dynamic interlocking mechanism with the displacement redundancy threshold of the S7 acceptance model; S7. Integrate S2 positioning topology data and S5 axial force cloud map to build a three-dimensional acceptance model, spatially align S2 positioning topology data with S5 axial force cloud map, and automatically identify non-compliant areas by comparing with design standards; For non-compliant areas, give priority to using S1 benchmark inclination correction strategy to add oblique support, and the second option is to trigger S6 grouting reinforcement mechanism; Iteratively optimize the S1 initial parameter group through machine learning algorithm to form a closed-loop control system with adaptive parameters.

2. A foundation pit oblique support construction method according to claim 1, characterized in that: The S1 specifically includes the following steps: S1-1. Depth logarithmic compensation: The natural logarithm of the foundation pit depth is used as the calculation base. When the depth reaches the 5-meter reference threshold, the pre-loaded axial force value is compensated at a gradient of 15% for every 1 unit increase in the natural logarithm. The maximum compensation value shall not exceed 200% of the design value, and the output parameter shall be used as the design basis for the pre-loaded axial force in step S5. S1-2. Geological compensation triggering. In the soil shear wave velocity detection, when the wave velocity values ​​of three consecutive detection points are lower than 80% of the critical threshold, the benchmark inclination compensation mechanism is triggered. The compensation amplitude is 5 degrees and the cumulative number of times does not exceed twice. The compensation result is synchronized in real time to the inclination control module of step S3.

3. A foundation pit oblique support construction method according to claim 1, characterized in that: The S2 specifically includes the following steps: S2-1. 3D grid construction, with the center of the bottom surface of the foundation pit as the coordinate origin, laser positioning points are arranged along the reference inclination direction according to the sinusoidal distribution density, the distance between adjacent points is inversely proportional to the square root of the foundation pit depth, and the distance is reduced by 20% for every 5 meters increase in depth, and the generated data is used as the reference input for weld angle compensation in step S4; S2-2. Dynamic tolerance adjustment: when the foundation pit depth exceeds 15 meters, the lateral deviation threshold is dynamically adjusted at a ratio of 0.3 mm per meter of depth, and the longitudinal deviation threshold is maintained at 120% of the initial value. Exceeding limit data directly triggers the acceptance model alarm mechanism of step S7.

4. A foundation pit oblique support construction method according to claim 1, characterized in that: The S3 specifically includes the following steps: S3-1. Window period calculation, establish an exponential decay relationship between excavation depth and installation window period, the initial 4-hour window period is shortened to 60% of the original length when the depth reaches 8 meters, the attenuation coefficient takes the natural logarithm of the depth value, and the output parameter controls the axial force collection frequency of step S5; S3-2. Safety time guarantee: by superimposing the minimum response time of the hydraulic system of 0.8 hours and the structural strain buffer time of 0.4 hours, the lower limit of the window period is determined to be 1.2 hours and cannot be exceeded. The limit data is transmitted back to the positioning verification system of step S2 in real time.

5. A foundation pit oblique support construction method according to claim 1, characterized in that: The S4 specifically comprises the following steps: S4-1. Multi-level torque control: Initial tightening stage: Apply 50% of the design torque ( ), maintain stress relaxation for 10 minutes; Secondary preload stage: Based on the weld angle compensation Q output by the dynamic adjustment model, according to the formula: Adjust the torque, where Q is calculated in real time by the S4 main step model; if the strain rate ε exceeds the reference value , then the additional strain compensation term is superimposed, and the torque upper limit is 150% of the design value. Final tightening stage: Apply 100% design torque ( ); S4-2. Residual stress verification: After the final tightening is completed, ultrasonic testing of residual stress distribution is used, and the difference between any two points is required to be ≤12%; When the limit is exceeded, the S5 axial force monitoring unit is called for cross-validation and the model error is calculated: By formula Reverse correction weight coefficient , the updated parameters are written into the dynamic adjustment model; The verification results are synchronized to the S7 acceptance model for subsequent construction section parameter optimization.

6. A foundation pit oblique support construction method according to claim 1, characterized in that: The S5 specifically includes the following steps: S5-1. Compensation trigger: when the real-time axial force value deviates from the design value by plus or minus 15%, the compensation mechanism linked to step S6 is activated. The compensation amount is calculated according to the square root ratio of the grouting pressure gradient, and the trigger threshold is dynamically adjusted according to the S1 pre-axial force reference value; S5-2. Unit reuse. When the support is removed, the monitoring unit is double-index calibrated, requiring the calibration error to be less than 0.8% and the protective structure to be intact. The reuse status data is encrypted and transmitted to the S7 acceptance system and fed back to the positioning accuracy optimization algorithm in step S2.

7. A foundation pit oblique support construction method according to claim 1, characterized in that: The S6 specifically comprises the following steps: S6-1. Pressure grading control: when the foundation displacement reaches the yellow warning zone of the thermal distribution map, 0.2MPa foundation grouting is started. After entering the red warning zone, the pressure is increased by 0.2MPa per level. The upper limit of pressure is 125% of the S7 displacement redundancy threshold. The warning grading standard is inherited from the S2 positioning deviation historical data; S6-2. Pressure interlock. When the grouting pressure gradient is increased to the third level, the S5 axial force data is automatically called to establish a pressure-axial force feedback loop. The pressure increase is adjusted to 80% of the previous value, and the feedback loop parameters are synchronized to the S3 excavation control timing table.

8. A foundation pit oblique support construction method according to claim 1, characterized in that: The S7 specifically comprises the following steps: S7-1. 3D model construction and deviation assessment: Based on the coordinate data of the S2 laser positioning grid (accuracy ±0.3mm) and the S5 axial force monitoring data (sampling frequency ≥1Hz), the following method is used to construct the 3D acceptance model: S7-1-1. The foundation pit support structure is meshed using the Delaunay triangulation algorithm. The mesh density varies with depth. ( , Dynamic encryption for depth; S7-1-2. Use Kriging interpolation method to generate continuous axial force cloud map, the calculation formula is: in, is the weight coefficient calculated for the semivariogram, is the drift term in the depth direction; Representing coordinates The predicted axial force value at ; Indicates the lateral coordinate of the foundation pit; represents the longitudinal coordinate of the foundation pit; z represents the vertical depth of the foundation pit; Indicates The measured axial force values ​​of the monitoring points are obtained through gridding collection of the wireless sensor units in step S5; S7-1-3. Establish a deviation assessment matrix: when It is judged as a non-compliant area; in: It indicates the lateral positioning deviation of the foundation pit support structure, that is, the difference between the actual position and the design coordinates in the lateral direction; Indicates longitudinal positioning deviation; represents the lateral deviation threshold; represents the longitudinal deviation threshold; Indicates the measured axial force value; Indicates the design axial force value; S7-2. Reinforcement decision driven by curvature analysis, cubic spline curvature analysis of non-compliant areas: S7-2-1. Fitting axial force distribution curve And calculate the curvature: in, Represents the curvature of the axial force distribution curve; : Axial force distribution function The first derivative of ; Represents the axial force distribution function The second derivative of S7-2-2. When When S1 reference inclination correction is triggered first in, represents the corrected reference inclination; Indicates the reference inclination before correction; : Sign function, the direction of inclination adjustment is determined according to the sign of the difference between the measured axial force and the designed axial force: If , , the inclination angle increases by 5°; if , , the inclination angle is reduced by 5° S7-2-3. When When S6 secondary grouting is triggered, the grouting pressure is: in, Indicates grouting pressure; Make dynamic adjustments; S7-3. Parameter adaptive optimization and verification: S7-3-1. Use the exponential moving average algorithm (EMA) to update the initial parameters of S1. The formula is: in, is the updated initial reference inclination, is the initial reference inclination before updating, is the actual reference inclination angle used in the current construction section, is the attenuation factor, is the weight of the current data, n represents the nth construction; S7-3-2. The updated parameters are simulated and verified by the S2 positioning system to ensure that the three-dimensional laser positioning deviation is ≤0.5mm / m (horizontally) and ≤1.0mm / m (vertically). After verification, they are synchronized to the construction control module; S7-3-3. Use support vector regression (SVR) to train historical data and optimize the weights of the deviation evaluation matrix, with positioning deviation accounting for 60% and axial force deviation accounting for 40%, forming a parameter optimization closed loop.

Images