A servo support foundation soil displacement evolution simulation method

Abstract

The present application relates to the technical field of geotechnical engineering and underground structure monitoring, in particular to a servo support foundation soil displacement evolution simulation method. The axial force, displacement and enclosure wall deformation sequence of servo loading are obtained; the overall stiffness loss coefficient of the soil is calculated by energy comparison; the damage distribution matrix is determined by deformation mode matching; the non-uniform stiffness correction field is generated and the numerical model is updated in combination with the two; the response is predicted based on the updated model and the servo control strategy is generated. The method realizes the quantification, positioning and modeling of hidden damage of the soil, and improves the accuracy and safety of foundation deformation control.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering and underground structure monitoring technology, specifically to a method for simulating the displacement evolution of soil in a servo-supported foundation pit. Background Technology

[0002] As urban underground space development deepens and expands, the safety and deformation control issues in deep foundation pit engineering are becoming increasingly prominent. Servo support (automatic axial force compensation) systems, due to their ability to actively and dynamically adjust the axial force of the supports to counteract the deformation of the retaining structure caused by foundation pit excavation, have become one of the mainstream technologies for controlling deep foundation pit deformation. This system monitors the displacement of the retaining structure in real time and adjusts the axial force of the hydraulic cylinders in reverse, forming a closed-loop control process of "monitoring-feedback-adjustment". In the field of deep foundation pit engineering, servo support systems are widely used to control the deformation of the retaining structure. This system counteracts the excavation unloading effect by applying active axial force to the supports. However, soil, as a nonlinear material, undergoes changes in its internal structure after experiencing repeated loading and unloading actions by the servo system.

[0003] Existing numerical simulation methods typically simplify soil into homogeneous material models. While this simplification is computationally convenient, it neglects the non-uniform structural damage that occurs in soft soil during servo loading. In particular, when the active axial force is applied to a certain extent, hidden shear strain concentration zones often form within the soil, leading to significant degradation of local stiffness. This "internal injury" cannot be represented by conventional homogeneous models. Therefore, existing models, unable to perceive this historical damage, often overestimate the soil's resistance to deformation, resulting in overly optimistic displacement control predictions and an inability to effectively identify potential engineering risks caused by localized shear failure. Summary of the Invention

[0004] To address the technical problem that existing models struggle to detect non-uniform structural damage within soil, overestimate soil's resistance to deformation, and result in overly optimistic displacement control predictions that fail to effectively identify potential engineering risks caused by localized shear failure, this invention aims to provide a servo-supported foundation pit soil displacement evolution simulation method. The specific technical solution adopted is as follows: This invention proposes a method for simulating the displacement evolution of soil in a servo-supported foundation pit, the method comprising: Acquire real-time axial force sequence, piston stroke sequence, soil parameters, and measured displacement sequence of the retaining wall during the loading process of the servo system; Based on the discrete time-series changes of the real-time axial force sequence and piston stroke sequence, the measured servo input energy is obtained; based on the distribution of the soil layer parameters and the measured displacement sequence of the retaining wall, the theoretical elastic potential energy of the foundation pit is calculated; based on the deviation between the measured servo input energy and the theoretical elastic potential energy, the overall stiffness loss coefficient of the foundation pit soil is calculated. Based on the standard deviation of the measured displacement sequence distribution of the retaining wall, the field distribution matrix is ​​determined; for each grid cell in the numerical model of the foundation pit, the shear influence range weight field is determined based on the mapping relationship between the field distribution matrix and the numerical model; based on the overall stiffness loss coefficient and the shear influence range weight field, the grid stiffness correction coefficient is determined. The numerical model is updated based on the mesh stiffness correction coefficient; the remaining servo loading process is simulated using the updated numerical model to determine the axial force-displacement response ratio; and a servo strategy is generated based on the axial force-displacement response ratio.

[0005] Furthermore, the soil layer parameters include: Within a preset elastic calibration window, the equivalent back wall earth pressure increment and displacement increment of each soil layer are collected synchronously; the soil layer parameters of each soil layer are obtained by ratio calculation using the equivalent back wall earth pressure increment as the numerator and the displacement increment as the denominator.

[0006] Furthermore, the method for calculating the measured servo input energy includes: In the real-time axial force sequence, the average axial force at each moment is determined based on the average of the axial force at each moment and the axial force at the previous moment; in the piston stroke sequence, the displacement increment at each moment is determined based on the difference between the piston stroke at each moment and the piston stroke at the previous moment; the mechanical work at each moment is determined based on the product of the average axial force and the displacement increment; the cumulative value of the mechanical work at all moments is calculated as the measured servo input energy.

[0007] Furthermore, the method for calculating the theoretical elastic potential energy includes: Obtain the soil layer thickness for each soil layer; for each soil layer, use the soil layer parameters as stiffness coefficients, combine the soil layer thickness of each soil layer with the measured displacement sequence of the retaining wall, and calculate the elastic potential energy of each soil layer using the elastic potential energy formula; sum up the elastic potential energy of all soil layers to obtain the theoretical elastic potential energy.

[0008] Furthermore, the method for calculating the overall stiffness loss coefficient includes: The difference between the measured servo input energy and the theoretical elastic potential energy is calculated as the dissipated energy; the dissipated energy is normalized using the measured servo input energy to obtain the overall stiffness loss coefficient.

[0009] Furthermore, the method for obtaining the on-site distribution matrix includes: Obtain the excavation depth of the foundation pit and the depth of the measuring points for each soil layer; based on the relative deviation between the excavation depth of the foundation pit and the depth of the measuring points for each soil layer, and the displacement distribution of the measured displacement sequence of the retaining wall, obtain the on-site deformation feature vector; take the sample corresponding to the maximum value of the cosine between the on-site deformation feature vector and the template displacement curve vector of each sample as the optimal matching term; extract the standard damage distribution matrix of the optimal matching term; map the standard damage distribution matrix to the on-site engineering coordinate system to obtain the on-site distribution matrix.

[0010] Furthermore, the method for obtaining the on-site deformation feature vector includes: The maximum value of the displacement of all soil layers at the current analysis time in the measured displacement sequence of the retaining wall is obtained as the reference displacement value. The displacement value at the current time is normalized according to the reference displacement value at the current time to obtain the relative displacement at the current time. The depth of the measuring point of each soil layer is normalized according to the excavation depth of the foundation pit to obtain the relative depth of each soil layer. Based on the relative displacement and relative depth, the on-site deformation feature vector is obtained.

[0011] Furthermore, the method for obtaining the weight field of the shear influence range includes: For each grid cell in the numerical model, the shortest distance between the grid cell and all damaged skeleton points in the field distribution matrix is ​​taken as the minimum skeleton distance; a Gaussian function is constructed, which is negatively correlated with the minimum skeleton distance; the shear influence weight is calculated based on the Gaussian function, and the shear influence range weight field is constructed based on the shear influence weights of all grid cells.

[0012] Furthermore, the method for calculating the mesh stiffness correction coefficient includes: For each grid cell in the numerical model, the product of the shear influence weight of the current grid cell, the overall stiffness loss coefficient, and the preset engineering adjustment factor is used as the grid stiffness correction coefficient.

[0013] Furthermore, the method for calculating the axial force-displacement response ratio includes: Obtain the predicted displacement sequence and the remaining axial force of the retaining wall; take the maximum displacement in the predicted displacement sequence as the numerator and the remaining axial force as the denominator, and obtain the axial force-displacement response ratio by ratio calculation.

[0014] The present invention has the following beneficial effects: This invention achieves objective quantification of the overall damage level of the soil by calculating the overall stiffness loss coefficient of the foundation pit soil, avoiding reliance on distorted survey parameters and accurately perceiving and quantifying the overall stiffness softening accumulated in the soil during loading history. By determining the field distribution matrix based on the similarity between the field deformation feature vector and the template displacement curve vector, it overcomes the limitation of traditional monitoring in perceiving internal soil damage and achieves spatial positioning of hidden shear damage zones within the soil. By updating the numerical model based on the calculated grid stiffness correction coefficient, it achieves dynamic reconstruction of the non-uniform damage state of the soil in the numerical model, overcoming the bottleneck of computational distortion caused by the inability of existing models to express local shear zones. By simulating the remaining servo loading process using the updated numerical model and generating servo strategies, it achieves a leap from passive response to active predictive control, effectively avoiding engineering risks induced by blind loading and improving the intelligence and safety level of foundation pit deformation control. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages 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.

[0016] Figure 1 This is a flowchart of a method for simulating the displacement evolution of soil in a servo-supported foundation pit, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a servo-supported foundation pit soil displacement evolution simulation method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of the servo-supported foundation pit soil displacement evolution simulation method provided by the present invention.

[0020] Please see Figure 1The diagram illustrates a flowchart of a servo-supported foundation pit soil displacement evolution simulation method according to an embodiment of the present invention. The method includes: S101: Obtain the real-time axial force sequence, piston stroke sequence, soil parameters, and measured displacement sequence of the retaining wall during the servo system loading process.

[0021] According to the principles of physics, work represents the physical quantity accumulated by force acting on an object in space, and is obtained by calculating the product of force and displacement in the direction of force. Therefore, in order to analyze the mechanical work done by the servo system on the foundation pit retaining structure and subsequently analyze the non-uniform structural damage inside the soil, this invention needs to use the axial force sensor built into the servo cylinder to obtain the real-time axial force sequence and the displacement sensor located on the piston rod to obtain the piston stroke sequence. The real-time axial force sequence records the applied axial force value of the servo cylinder over time; the piston stroke sequence records the extension and retraction displacement of the servo cylinder piston rod over time. At the same time, before analyzing the impact of the servo system on the internal structure of the soil, it is necessary to obtain soil layer parameters to characterize the initial physical and mechanical properties of the soil on site, thereby providing a theoretical initial reference benchmark for the soil state undisturbed by construction. Because the spatially non-uniformly distributed shear damage (such as strain concentration zones) induced by the servo loading history inside the soil will significantly change the local support stiffness of the soil on which the retaining wall is based. This non-uniform change in internal stiffness will inevitably be directly reflected in the deformation of its external support, namely the retaining wall. Therefore, the measured horizontal displacement of the retaining wall along the depth direction is obtained and arranged in ascending order of depth to construct the measured displacement sequence of the retaining wall. The measured displacement sequence of the retaining wall includes the measured horizontal displacement values ​​corresponding to different depths of the retaining wall.

[0022] Preferably, in some possible implementations of the embodiments of the present invention, the soil layer parameters include: Within a preset elastic calibration window, the equivalent back wall earth pressure increment and displacement increment of each soil layer are collected synchronously; the soil layer parameters of each soil layer are obtained by ratio calculation using the equivalent back wall earth pressure increment as the numerator and the displacement increment as the denominator.

[0023] It should be noted that the system collects the axial force values ​​at each soil layer depth within the preset elastic calibration window; by subtracting the axial force value at the start time from the axial force value at the end time of the preset elastic calibration window, the axial force increment is obtained; based on the axial force increment, the equivalent earth pressure increment behind the wall is obtained using the internal force calculation method of the elastic foundation beam; and the horizontal displacement change at each soil layer depth within the preset elastic calibration window is recorded using the retaining wall inclinometer as the displacement increment.

[0024] It should be noted that the method for calculating the internal forces of beams on elastic foundations is a well-known technique in the field, and its calculation process will not be elaborated here.

[0025] In one specific implementation of this invention, since obtaining soil layer parameters is to characterize the initial physical and mechanical properties of the soil on site, and the destruction of the soil's internal structure by the servo system is a slow and continuous process, in order to obtain a theoretical soil state that is not disturbed by construction, it is necessary to set a very small load threshold as the calibration cutoff point. In this embodiment of the invention, the calibration cutoff point is set to 2% of the total axial force increment of the servo target, and the time period from the start of the servo system loading to the calibration cutoff point is set as a preset elastic calibration window.

[0026] S102: Based on the discrete time-series changes of the real-time axial force sequence and piston stroke sequence, obtain the measured servo input energy; based on the distribution of the soil layer parameters and the measured displacement sequence of the retaining wall, calculate the theoretical elastic potential energy of the foundation pit; based on the deviation between the measured servo input energy and the theoretical elastic potential energy, calculate the overall stiffness loss coefficient of the foundation pit soil.

[0027] During the movement of the servo cylinder, the product of its thrust and displacement represents the mechanical work done on the foundation pit system. Therefore, by integrating the axial force value at each moment in the real-time axial force sequence and the displacement value at each moment in the piston stroke sequence, the total energy input to the servo system since the start of loading can be accurately accumulated, serving as the sole objective benchmark for subsequent assessment of whether the soil has consumed additional energy due to damage. Simultaneously, since the soil layer parameters characterize the initial physical and mechanical properties of each soil layer on site, according to the elastic foundation beam theory, the stress and deformation of the soil structure satisfy a linear relationship. Therefore, based on the distribution characteristics of the soil layer parameters, the theoretical elastic potential energy that should be stored under the current retaining wall displacement state, assuming the soil is completely elastic and undamaged, can be calculated; this value represents the maximum energy storage capacity of the soil under ideal conditions. Considering that the difference between the measured servo input energy and the theoretical elastic potential energy reflects to some extent the energy irreversibly consumed by damage mechanisms such as the development of internal cracks in the soil and particle friction slip, the overall stiffness loss coefficient of the foundation pit soil is calculated based on this characteristic. The larger the coefficient, the more severe the overall softening of the soil and the more serious the damage, thus realizing a direct numerical characterization of the degree of damage.

[0028] It should be noted that, in one specific implementation of the present invention, the 2%-15% stage of the total axial force increment of the servo target is taken as the stage for obtaining the overall stiffness loss coefficient in step S102.

[0029] Preferably, in some possible implementations of the embodiments of the present invention, the method for calculating the measured servo input energy includes: The servo system's work on the foundation pit is a continuous process, where force and displacement change continuously over time. The total work is physically the area under the force-displacement curve. However, in reality, only sampled data at discrete time points can be obtained. Therefore, in this invention, a numerical method is used to approximately calculate this integral based on the real-time axial force sequence and the piston stroke sequence.

[0030] Specifically, the current moment and the previous moment are taken as adjacent moments. The average value of the axial forces at adjacent moments is calculated according to the real-time axial force sequence as the average axial force; the difference in piston strokes at adjacent moments is calculated according to the piston stroke sequence as the displacement increment; the product of the average axial force and the displacement increment is calculated as the mechanical work at adjacent moments, and by accumulating the mechanical work at all adjacent moments, the measured servo input energy is obtained.

[0031] As an example, in a specific implementation manner of an embodiment of this invention, the measured servo input energy The calculation formula can be expressed as: Where, represents the sampling moment index, ; represents the total number of sampling moments, forming continuous sampling intervals. Among them, when it represents the initial moment of the acquisition stage of the overall stiffness reduction coefficient set in step S102; represents the axial force value corresponding to the sampling moment index in the real-time axial force sequence; represents the axial force value corresponding to the sampling moment index in the real-time axial force sequence; represents the piston stroke corresponding to the sampling moment index in the piston stroke sequence; represents the piston stroke corresponding to the sampling moment index in the piston stroke sequence; In the servo system, the axial force changes linearly within each sampling interval. Therefore, the average value of the axial forces at two adjacent moments is used as the average axial force within this time interval, that is represents the average axial force within the time interval from the sampling moment index to the sampling moment index; and represents the displacement increment generated due to the axial force within the time interval from the sampling moment index to the sampling moment index, which is caused by the axial force at the sampling moment index time and the The difference in piston stroke between two sampling time points is obtained; according to the principles of physics, the mechanical work done during the time interval between two adjacent time points is the product of the average axial force and the displacement increment generated during that time interval, i.e. By accumulating the mechanical work over all time intervals, the measured servo input energy of the servo system during the entire sampling phase can be obtained.

[0032] Preferably, in some possible implementations of the embodiments of the present invention, the theoretical elastic potential energy calculation method includes: Since the soil layer parameters characterize the initial physical and mechanical properties of each soil layer on site, according to the elastic foundation beam theory, the stress and deformation of the soil structure satisfy a linear relationship. Therefore, based on the distribution characteristics of the soil layer parameters, the theoretical elastic potential energy that should be stored under the current retaining wall displacement state, if the soil is completely elastic and undamaged, can be calculated. The physical meaning of this value is the upper limit of energy absorption of the soil under ideal conditions.

[0033] Specifically, the thickness of each soil layer is obtained; for each soil layer, the soil layer parameters are used as stiffness coefficients, and the elastic potential energy of each soil layer is calculated using the elastic potential energy formula in combination with the soil layer thickness and the measured displacement sequence of the retaining wall; the elastic potential energy of all soil layers is accumulated to obtain the theoretical elastic potential energy.

[0034] As an example, in one specific implementation of this invention, the theoretical elastic potential energy... The calculation formula can be expressed as: in, Indicates the first Soil layer parameters; Indicates the first [number]th ... Displacement values ​​of soil layers; Indicates the first Soil layer thickness; according to the linear elastic potential energy formula in physics: for a spring constant of... The spring, causing it to deform. The stored elastic potential energy is This invention uses soil layer parameters as stiffness coefficients and the measured displacement values ​​in the measured displacement sequence of the retaining wall as the displacement of the soil structure deformation. Calculate the first Elastic potential energy density of soil per unit area; Through the first The elastic potential energy density of soil per unit area and the first The product of the soil layer thicknesses represents the product of the soil layers. The elastic potential energy stored in the soil layers; finally, by summing up the elastic potential energy of all soil layers, the theoretical elastic potential energy of the entire foundation pit soil can be obtained.

[0035] It should be noted that, in one specific implementation of this invention, the soil layer thickness refers to the vertical height range controlled by the inclinometer measuring point, which can be obtained from the inclinometer monitoring scheme document.

[0036] Preferably, in some possible implementations of the embodiments of the present invention, the method for calculating the overall stiffness loss coefficient includes: The difference between the measured servo input energy and the theoretical elastic potential energy is calculated as dissipated energy, representing the irrecoverable energy consumed by soil damage. The dissipated energy is normalized using the measured servo input energy to obtain the overall stiffness loss coefficient.

[0037] As an example, in one specific implementation of this invention, the overall stiffness loss coefficient... The calculation formula can be expressed as: in, This represents the measured servo input energy, quantifying the total energy input by the servo system to the foundation pit soil; It represents the theoretical elastic potential energy, which quantifies the upper limit of energy absorption of soil under ideal conditions; This is a preset minimum positive number to prevent the denominator from being zero; To find the maximum value of the function. Numerator term This represents dissipated energy, calculated by the difference between the measured servo input energy and the theoretical elastic potential energy. It represents the energy consumed by damage behaviors such as the propagation of internal soil cracks and particle slippage. This dissipated energy directly corresponds to the accumulation of soil degradation and damage in the foundation pit; The denominator term... The total energy reference used for normalization aims to obtain a dimensionless scaling factor. The measured servo input energy quantifies the total energy input by the servo system to the foundation pit soil; therefore, it is used... Normalization is performed using this as the denominator. Considering that in practical engineering, extreme cases may occur due to measurement noise, model errors, or calculation errors, causing the calculated theoretical elastic potential energy to be slightly greater than the measured servo input energy, resulting in a negative numerator. Physically speaking, dissipated energy cannot be negative (soil cannot generate energy out of thin air), therefore, normalization is performed using... When the numerator is negative, the overall stiffness reduction factor is forcibly set to 0, which is equivalent to determining that no obvious energy dissipation is detected under the current working condition, that is, it is considered that no observable overall damage has occurred in the soil and the stiffness has not been reduced.

[0038] It should be noted that the range of the overall stiffness loss coefficient is [value missing]. When the overall stiffness loss coefficient is 0, it indicates that the soil is basically in an elastic state and the damage is minor; when the overall stiffness loss coefficient is close to 1, it indicates that the soil is severely damaged, almost all of the input energy is dissipated, and the stiffness degradation is significant.

[0039] In one specific implementation of this invention, during the initial loading phase of the servo system, the measured servo input energy may be extremely small, close to zero. Therefore, to avoid the denominator being zero, a preset minimum positive number is required. Furthermore, considering that the unit of the measured servo input energy is J, the preset minimum positive number is set to... J, thereby eliminating the influence of dimensions.

[0040] S103: Determine the field distribution matrix based on the standard deviation of the measured displacement sequence distribution of the retaining wall; for each grid cell in the numerical model of the foundation pit, determine the shear influence range weight field based on the mapping relationship between the field distribution matrix and the numerical model; determine the grid stiffness correction coefficient based on the overall stiffness loss coefficient and the shear influence range weight field. Since the location of damage cannot be directly observed, but the spatially non-uniformly distributed shear damage (such as strain concentration zones) induced by the servo loading history within the soil significantly alters the local support stiffness of the soil upon which the retaining wall rests. This non-uniform change in internal stiffness will inevitably be directly reflected in the deformation morphology of its external support, i.e., the retaining wall. Therefore, this invention determines the field distribution matrix based on the standard deviation of the measured displacement sequence distribution of the retaining wall. This matrix is ​​the skeleton diagram of the most likely damage location within the soil, derived from the current macroscopic deformation. To transform the discrete damage location skeleton into a continuously varying influence intensity field in the numerical model space, and further determine the shear influence range weight field based on the mapping relationship between the field distribution matrix and the numerical model.

[0041] Considering that the structural damage caused by the servo system to the soil is non-uniform, and that the shear influence range weight field characterizes the influence of the overall stiffness loss coefficient in each grid cell of the numerical model, the grid stiffness correction coefficient can be determined based on the overall stiffness loss coefficient and the shear influence range weight field to characterize the damage level of each grid cell. In areas with severe damage (high overall stiffness loss coefficient) and located in the core of the shear zone (high shear influence range weight), the grid stiffness correction coefficient is large; in areas with slight damage or far from the shear zone, the grid stiffness correction coefficient is small or even zero.

[0042] It should be noted that in the field of deep foundation pit engineering design, analysis, and safety assessment, finite element models are commonly used for analysis because they can be seamlessly integrated with existing engineering practices. Therefore, this embodiment of the invention uses a finite element model as the numerical model. The finite element model is a well-known technique among those skilled in the art, and will not be further limited or elaborated upon here.

[0043] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the field distribution matrix includes: Obtain the excavation depth of the foundation pit and the depth of the measuring points at each moment; based on the relative deviation between the excavation depth and the measuring point depth of each soil layer, and the displacement distribution of the measured displacement sequence of the retaining wall, obtain the on-site deformation feature vector; take the sample corresponding to the maximum value of the cosine between the on-site deformation feature vector and the template displacement curve vector of each sample as the optimal matching item, wherein the range of the cosine value is... The closer the cosine value is to 1, the more consistent the direction of the on-site deformation feature vector and the template displacement curve vector, meaning the higher the similarity between the deformation shape of the on-site retaining wall and the deformation shape of the template sample. The standard damage distribution matrix of the optimal matching term is extracted; this matrix represents the damage distribution that best matches the current foundation pit soil. Since the standard damage distribution matrix is ​​a dimensionless, relative coordinate-based damage location template extracted from model tests or standard numerical simulations, it lacks any actual engineering dimensions and coordinates. What needs to be updated subsequently is a numerical model of the foundation pit with real dimensions and located in a specific geographical location. Therefore, the standard damage distribution matrix needs to be mapped to the on-site engineering coordinate system to obtain the on-site distribution matrix.

[0044] In one specific implementation of this invention, the method for obtaining the field distribution matrix includes: Define the mapping area: with the top of the enclosure wall as the origin (0,0), this area covers a horizontal distance of... And the range of depth ,in The excavation depth of the foundation pit refers to the vertical distance from the bottom of the pit to the ground surface, which can be obtained from the design drawings or the on-site construction progress.

[0045] Perform mapping: For each grid cell in the numerical model that lies within the mapping region. Calculate its relative coordinates Based on the relative coordinates, find the corresponding pixel value (0 or 1) in the standard damage distribution matrix and assign it to the field distribution matrix. For grid cells located outside the mapping region, assign values ​​directly. In the field distribution matrix, a grid cell with a value of 1 indicates that the soil structure of that grid cell has been significantly damaged by the servo system; a grid cell with a value of 0 indicates that the soil structure of that grid cell has been slightly damaged by the servo system.

[0046] Furthermore, in a specific implementation of this invention, the method for obtaining the template displacement curve vector and its standard damage distribution matrix for each sample includes: Since the deformation modes of foundation pit retaining walls are diverse, this embodiment of the invention targets the typical deformation modes of foundation pit retaining walls such as cantilever, convex, and skirting types, and obtains the corresponding soil shear strain rate field images through particle image velocimetry technology.

[0047] For each soil shear strain rate field image, the maximum flow minimum cut algorithm is used to identify the banded regions of concentrated shear strain, which are then binarized. The spatial location of these regions is marked as 1, while the remaining background regions are marked as 0, generating a dimensionless standard damage distribution matrix. This matrix describes the relative locations of damage occurring within the soil under specific deformation modes.

[0048] For each type of foundation pit retaining wall deformation mode, the simulated displacement sequence of the retaining wall corresponding to each template sample is extracted. Based on the principle of obtaining the on-site deformation feature vector, the simulated displacement sequence of the retaining wall is normalized to obtain the template displacement curve vector of each sample.

[0049] It should be noted that particle image velocimetry and the maximum flow minimum cut algorithm are both well-known techniques in the field, and will not be elaborated upon here.

[0050] It should be noted that, in one specific implementation of this invention, the template samples calibrated by the system cannot exhaustively represent all possible and ever-changing foundation pit deformation modes in reality. Due to unknown geological conditions, special construction disturbances, unforeseen loads, or structural defects, a completely new and previously unrecorded deformation pattern may occur on-site. Therefore, by setting a similarity threshold, if the maximum value of the cosine of the on-site deformation feature vector and the template displacement curve vector of each sample is less than the similarity threshold, it indicates that the system has not collected the corresponding template sample for the current foundation pit deformation mode, and the system triggers an alarm: "The foundation pit deformation is abnormal and may exceed conventional understanding; high vigilance and manual intervention are required." Simultaneously, using only the overall stiffness loss coefficient obtained in step S102, based on the principle of obtaining the on-site distribution matrix, it is uniformly mapped onto each grid cell of the numerical model, thereby obtaining the on-site distribution matrix.

[0051] In one specific implementation of this invention, based on engineering reliability, the similarity threshold is set to 0.7. This means that the angle between the on-site deformation feature vector and the template displacement curve vector is approximately 45°, and their directions are significantly consistent. This effectively filters out low-quality matches that appear plausible but are not, ensuring that the matching results have sufficient credibility to guide model updates; it also allows for a greater chance of identifying real and typical deformation patterns, enabling the system's main functions to be realized.

[0052] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the on-site deformation feature vector includes: Considering that even within the same foundation pit, although the degree of displacement may differ at different loading stages, the shape of the retaining wall's displacement curve (e.g., the location of the bulge) may be consistent, directly comparing the displacement curves of the retaining wall at each loading stage would obscure the similarity in shape due to numerical differences. Therefore, by obtaining the maximum displacement value of all soil layer depths in the measured displacement sequence of the retaining wall at the current analysis time as the benchmark displacement value, the displacement value at the current analysis time is normalized based on the benchmark displacement value to obtain the relative displacement at the current analysis time. This preserves the relative undulation of the measured displacement curve of the retaining wall along the depth while filtering out absolute magnitudes. Since the displacement curve shapes of the retaining walls at different foundation pit depths may be similar, but their relative depth coordinates are completely different, a measuring point is set for each soil layer. The soil layer depth of each measuring point is obtained according to the inclinometer monitoring scheme file, and the measuring point depth of each soil layer is normalized according to the foundation pit excavation depth to obtain the relative depth of each soil layer, thereby eliminating the influence of differences in the depth scale of different foundation pits. Based on the relative displacement and relative depth, the on-site deformation feature vector is obtained.

[0053] In one specific implementation of this invention, the displacement value at the current moment is used as the numerator, and the reference displacement value at the current moment is used as the denominator. These values ​​are then normalized using a ratio operation to determine the relative displacement at the current moment. Similarly, the depth of each soil layer's measuring point is used as the numerator, and the excavation depth of the foundation pit is used as the denominator. These values ​​are then normalized using a ratio operation to determine the relative depth of each soil layer. The relative displacements at the current analysis moment are arranged in ascending order along the relative depths of each soil layer to obtain the field deformation feature vector. This field deformation feature vector is a sequence containing the relative displacements of the soil layers at each relative depth.

[0054] It should be noted that, in this embodiment of the invention, the cutoff time of the 2%-15% stage of the total axial force increment of the servo target is taken as the current analysis time.

[0055] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the shear influence range weight field includes: For each grid cell in the numerical model, a point with a value of 1 in the field distribution matrix is ​​defined as a damaged skeleton point. The shortest distance between the grid cell and all damaged skeleton points in the field distribution matrix is ​​taken as the minimum skeleton distance. The smaller this value, the closer the current grid cell is to the damaged skeleton point, and the greater the shearing influence weight it receives from the servo system. A Gaussian function is constructed, which is negatively correlated with the minimum skeleton distance. The shearing influence weight of each grid cell is calculated based on the Gaussian function, and the shearing influence weight is used as the label value of each grid cell to obtain a shearing influence range weight field that describes the shearing influence weight of all grid cells.

[0056] As an example, in one specific implementation of this invention, the mesh cell Shearing effect weight The calculation formula can be expressed as: in, This represents the minimum skeleton spacing, which mathematically means the mesh cell. The shortest Euclidean distance to all damaged skeletal points; This indicates the preset characteristic width. Because the influence of shear damage zones (such as shear strain concentration zones) in the soil on the mechanical properties (such as stiffness) of the surrounding soil does not abruptly stop at the boundary, but gradually weakens and smoothly decays with increasing distance from the core zone. This decay stems from the diffusion of the stress field and the continuity of material deformation. The Gaussian function... The function reaches a maximum value of 1 at the center point (x=0), and then smoothly, continuously, and monotonically decays to 0 as |x| increases. Its decay pattern (exponential decay that starts fast and then slows down) conforms to the characteristic in engineering that nearby effects are severe while distant effects are weak, providing the closest mathematical description of the decay of influence with distance. Therefore, this invention constructs a Gaussian function for quantifying the shearing of influence weights based on the minimum skeleton distance and a preset feature width. The weight of shear effect is negatively correlated with the minimum skeleton distance.

[0057] It should be noted that, in one specific implementation of this invention, to ensure the stability of the numerical simulation, if the preset feature width is too small, the damage band will be too narrow and cannot be captured by the mesh; if the preset feature width is too large, the damage range will spread unreasonably. Therefore, in this embodiment of the invention, the preset feature width is set to 1.5 times the side length of the numerical model mesh cell, which can be adjusted according to the specific implementation scenario.

[0058] Preferably, in some possible implementations of the embodiments of the present invention, the method for calculating the mesh stiffness correction coefficient includes: Since the overall stiffness loss coefficient quantitatively diagnoses the average damage level of the soil from a macroscopic perspective, while the shear influence weight characterizes the influence of the diagnosed damage zone on each grid cell, calculating the product of the two can achieve non-uniform distribution of the damage level according to the influence of the damage zone on each grid cell. Finally, by introducing a preset engineering adjustment factor, the theoretical algorithm is moved to engineering application.

[0059] Specifically, for each grid cell in the numerical model, the product of the shear influence weight of the current grid cell, the overall stiffness loss coefficient, and the preset engineering adjustment factor is used as the grid stiffness correction coefficient.

[0060] It should be noted that in one specific implementation of this invention, since the preset engineering adjustment factor is used to fine-tune the reduction sensitivity based on the historical fitting accuracy, the value range is [1, 1.5]. In this embodiment of the invention, it is taken as 1.2, which can be adjusted according to the specific implementation scenario.

[0061] S104: Update the numerical model according to the mesh stiffness correction coefficient; use the updated numerical model to simulate the remaining servo loading process and determine the axial force-displacement response ratio; generate a servo strategy based on the axial force-displacement response ratio.

[0062] The mesh stiffness correction coefficient calculated in step S103 directly quantifies the correction value of the elastic modulus of each soil mesh element. Therefore, by updating the numerical model using the mesh stiffness correction coefficient, a numerical model characterizing the current soil damage state can be obtained. The updated numerical model is used to simulate the remaining servo loading process, determining the axial force-displacement response ratio, thereby quantifying the displacement rebound gain obtained per unit axial force input, reflecting the soil damage state. Finally, a servo strategy is generated based on the axial force-displacement response ratio to provide early warning of foundation pit deformation.

[0063] It should be noted that, in a specific implementation of this invention, the remaining servo loading process is the end time of step S102, and the incomplete servo loading process corresponds to the 15%-100% stage of the total axial force increment of the servo target.

[0064] In one specific implementation of this invention, the numerical model update process includes: in, Represents grid cells The updated elastic modulus is used to characterize the soil stiffness of this mesh element; This represents the initial elastic modulus of the soil, i.e., its stiffness when undamaged (which can be obtained from geotechnical engineering investigation documents at the initial stage of construction). Represents grid cells The mesh stiffness correction factor; This indicates the preset residual modulus threshold. This indicates the reduction ratio. At the center of the shear band (where the mesh stiffness correction factor is largest, close to 1), the reduction ratio is largest, and the modulus reduction is greatest; far from the shear band (where the mesh stiffness correction factor is smallest, close to 1), the reduction ratio is largest. The modulus remains unchanged. This indicates that the initial elastic modulus is reduced non-uniformly based on the reduction ratio; The function ensures that even in the most severely damaged regions, the element modulus does not drop to 0 or a negative value. This is to prevent the finite element stiffness matrix from becoming singular and to prevent computational non-convergence.

[0065] It should be noted that, in one specific implementation of this invention, the servo strategy includes: Based on the axial force-displacement response ratio and a preset performance threshold (set to be [value] in this embodiment of the invention) The system compares the axial force-displacement response ratio with the preset efficiency threshold, automatically generating control strategy commands based on the comparison results. When the axial force-displacement response ratio is greater than or equal to the preset efficiency threshold, it indicates that the soil stiffness is still good and the servo loading efficiency is in the high-efficiency zone, outputting "Continue to execute the original servo loading scheme." The system allows the servo cylinder to continue increasing the axial force as planned. When the axial force-displacement response ratio is less than the preset efficiency threshold, it indicates that the soil has undergone significant shear damage, and the servo loading efficiency has entered the low-efficiency zone or failure zone. At this time, continuing to increase the axial force will not only fail to significantly reduce the displacement, but may also induce local instability of the foundation pit; outputting "Soil damage warning," the system automatically cuts off the pressurization command and sends a displacement over-limit warning signal to the monitoring center.

[0066] Preferably, in some possible implementations of the embodiments of the present invention, the method for calculating the axial force-displacement response ratio includes: Obtain the predicted displacement sequence and the remaining axial force of the retaining wall; take the maximum displacement in the predicted displacement sequence as the numerator and the remaining axial force as the denominator, and obtain the axial force-displacement response ratio by ratio calculation.

[0067] In one specific implementation of this invention, the remaining axial force to be loaded by the servo system is determined. Its value is the total target axial force set by the servo control scheme minus the axial force applied at the cutoff time of step S102. In the updated numerical model, based on the principle of obtaining the measured displacement sequence of the retaining wall in step S101, the predicted displacement sequence of the retaining wall is obtained, and the maximum displacement is extracted as the numerator, representing the expected effect of servo loading. The larger this value, the better the control effect of the servo system. The remaining axial force to be loaded is used as the denominator, representing the cost invested in servo loading, i.e., the axial force increment. The axial force-displacement response ratio is obtained through ratio calculation. A higher axial force-displacement response ratio indicates that the soil is in the elastic stage, and the axial force can be efficiently converted into rebound; a lower value indicates that the soil has entered plastic flow, and the axial force is mainly consumed by internal shear slip.

[0068] It should be noted that, in one specific implementation of this invention, considering that the denominator may approach zero when the remaining axial force to be loaded is close to zero (e.g., when loading is nearing completion), a preset minimum force value is introduced to ensure the stability of the division operation and prevent program errors. In this embodiment, the preset minimum force value is set to 0.1 kN, which can be adjusted according to the specific implementation scenario.

[0069] In summary, this invention achieves objective quantification of the overall damage level of the soil by calculating the overall stiffness loss coefficient of the foundation pit soil, avoiding reliance on distorted survey parameters and accurately perceiving and quantifying the overall stiffness softening accumulated in the soil during loading history. By determining the field distribution matrix based on the similarity between the field deformation feature vector and the template displacement curve vector, it overcomes the limitation of traditional monitoring in perceiving internal soil damage and achieves spatial positioning of hidden shear damage zones within the soil. By updating the numerical model based on the calculated grid stiffness correction coefficient, it achieves dynamic reconstruction of the non-uniform damage state of the soil in the numerical model, overcoming the bottleneck of computational distortion caused by the inability of existing models to express local shear zones. By simulating the remaining servo loading process using the updated numerical model and generating servo strategies, it achieves a leap from passive response to active predictive control, effectively avoiding engineering risks induced by blind loading and improving the intelligence and safety level of foundation pit deformation control.

[0070] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0071] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for simulating the displacement evolution of soil in a servo-supported foundation pit, characterized in that, The method includes: Acquire the real-time axial force sequence, piston stroke sequence, soil parameters, and measured displacement sequence of the retaining wall during the loading process of the servo system; Based on the discrete time-series changes of the real-time axial force sequence and piston stroke sequence, the measured servo input energy is obtained; based on the distribution of the soil layer parameters and the measured displacement sequence of the retaining wall, the theoretical elastic potential energy of the foundation pit is calculated; based on the deviation between the measured servo input energy and the theoretical elastic potential energy, the overall stiffness loss coefficient of the foundation pit soil is calculated. Based on the standard deviation of the measured displacement sequence distribution of the retaining wall, the field distribution matrix is ​​determined; for each grid cell in the numerical model of the foundation pit, the shear influence range weight field is determined based on the mapping relationship between the field distribution matrix and the numerical model; based on the overall stiffness loss coefficient and the shear influence range weight field, the grid stiffness correction coefficient is determined. The numerical model is updated according to the mesh stiffness correction coefficient; the remaining servo loading process is simulated using the updated numerical model to determine the axial force-displacement response ratio; a servo strategy is generated based on the axial force-displacement response ratio. The method for calculating the overall stiffness loss coefficient includes: The difference between the measured servo input energy and the theoretical elastic potential energy is calculated as the dissipated energy; the dissipated energy is normalized using the measured servo input energy to obtain the overall stiffness loss coefficient. The method for obtaining the on-site distribution matrix includes: Obtain the excavation depth of the foundation pit and the depth of each soil layer's measuring points; based on the relative deviation between the excavation depth and the measuring point depths of each soil layer, and the displacement distribution of the measured displacement sequence of the retaining wall, obtain the on-site deformation feature vector; take the sample corresponding to the maximum cosine value between the on-site deformation feature vector and the template displacement curve vector of each sample as the optimal matching term; extract the standard damage distribution matrix of the optimal matching term; map the standard damage distribution matrix to the on-site engineering coordinate system to obtain the on-site distribution matrix; The method for obtaining the on-site deformation feature vector includes: The maximum value of the displacement of all soil layers at the current analysis time in the measured displacement sequence of the retaining wall is obtained as the reference displacement value. The displacement value at the current time is normalized according to the reference displacement value to obtain the relative displacement at the current time. The depth of the measuring point of each soil layer is normalized according to the excavation depth of the foundation pit to obtain the relative depth of each soil layer. Based on the relative displacement and relative depth, the on-site deformation feature vector is obtained. The method for obtaining the weighted field of the shear influence range includes: For each grid cell in the numerical model, the shortest distance between the grid cell and all damaged skeleton points in the field distribution matrix is ​​taken as the minimum skeleton distance; a Gaussian function is constructed, which is negatively correlated with the minimum skeleton distance; the shear influence weight is calculated based on the Gaussian function, and the shear influence range weight field is constructed based on the shear influence weight of all grid cells; The method for calculating the mesh stiffness correction coefficient includes: For each grid cell in the numerical model, the product of the shear influence weight of the current grid cell, the overall stiffness loss coefficient, and the preset engineering adjustment factor is used as the grid stiffness correction coefficient.

2. The servo-supported foundation pit soil displacement evolution simulation method according to claim 1, characterized in that, The soil layer parameters include: Within a preset elastic calibration window, the equivalent back wall earth pressure increment and displacement increment of each soil layer are collected synchronously; the soil layer parameters of each soil layer are obtained by ratio calculation using the equivalent back wall earth pressure increment as the numerator and the displacement increment as the denominator.

3. The servo-supported foundation pit soil displacement evolution simulation method according to claim 1, characterized in that, The method for calculating the measured servo input energy includes: In the real-time axial force sequence, the average axial force at each moment is determined based on the average of the axial force at each moment and the axial force at the previous moment; in the piston stroke sequence, the displacement increment at each moment is determined based on the difference between the piston stroke at each moment and the piston stroke at the previous moment; the mechanical work at each moment is determined based on the product of the average axial force and the displacement increment; the cumulative value of the mechanical work at all moments is calculated as the measured servo input energy.

4. The servo-supported foundation pit soil displacement evolution simulation method according to claim 1, characterized in that, The theoretical elastic potential energy calculation method includes: Obtain the soil layer thickness for each soil layer; for each soil layer, use the soil layer parameters as stiffness coefficients, combine the soil layer thickness of each soil layer with the measured displacement sequence of the retaining wall, and calculate the elastic potential energy of each soil layer using the elastic potential energy formula; sum up the elastic potential energy of all soil layers to obtain the theoretical elastic potential energy.

5. The servo-supported foundation pit soil displacement evolution simulation method according to claim 1, characterized in that, The method for calculating the axial force-displacement response ratio includes: Obtain the predicted displacement sequence and the remaining axial force of the retaining wall; take the maximum displacement in the predicted displacement sequence as the numerator and the remaining axial force as the denominator, and obtain the axial force-displacement response ratio by ratio calculation.

Images