A high-pressure jet grouting pile construction optimization control method

Abstract

The application provides a high-pressure rotary jet pile construction optimization control method, and relates to the technical field of rock foundation construction control.The application obtains soil types and permeability coefficients layer by layer along the depth, constructs a slurry diffusion model based on fluid-solid coupling theory, calculates a pressure field, and outputs a slurry saturation space distribution matrix.A strength mapping program is used to obtain a pile body compressive strength prediction value, a gradient descent correction mechanism is activated after comparing the prediction value with a target strength, the correction amplitude is adaptively adjusted according to the soil types, and the boundary initial value is updated.After convergence, the optimal slurry flow and jetting pressure are used as the boundary initial value of the next depth layer, and the optimal construction parameter sequence of the whole pile length is generated by layer-by-layer recursion to the pile bottom.The interlayer transfer of the slurry flow and the jetting pressure guarantees the physical continuity of the construction state, and the gradient correction amplitude is adaptively matched with the stratum condition, so that the optimization result directly corresponds to the equipment operation freedom, and the control accuracy and the executability are improved.

Description

Technical Field

[0001] This invention relates to the field of construction control technology for rock and soil foundations, specifically to an optimized control method for high-pressure jet grouting pile construction. Background Technology

[0002] High-pressure jet grouting is a foundation treatment process that uses a high-pressure jet to cut and mix the soil and inject solidifying grout, thereby forming a columnar reinforced body with specific strength underground. Because the construction process takes place entirely underground, the pile's formation and strength distribution cannot be directly observed during construction. Therefore, accurately predicting the pile strength before or during construction and generating reliable construction control parameters accordingly is a crucial technical aspect for ensuring pile quality and project safety.

[0003] A common technical approach for optimizing and controlling high-pressure jet grouting pile construction is to collect historical pile strength data covering different soil types and combinations of construction parameters, construct a data-driven prediction model with jetting pressure, grout flow rate, drill rod lifting speed, and drill rod rotation speed as inputs and pile strength as output, and then use swarm intelligence optimization algorithms to search for the optimal parameter combination in the construction parameter space that makes the predicted strength approximate the target strength. The core assumption of this approach is that there exists a static mapping relationship between construction parameters and the final strength that can be fully learned from historical data. However, the above-mentioned technical approach has at least the following two shortcomings in practical applications, and these shortcomings cannot be solved within its existing technical framework. The existing technical framework treats each depth layer as an isolated independent unit for parameter optimization, and cannot incorporate the continuous transmission relationship of construction state between adjacent depth layers into its prediction and optimization model. Due to the lack of a physical state transmission mechanism between layers in the depth direction, the existing technology generates a physical discontinuity in the full-length construction parameter sequence at the layer junction. When implementing control based on this parameter sequence, it is very easy for strength abrupt changes or weak interlayers to occur near the soil interface due to parameter abrupt changes. This defect cannot be compensated for by increasing the amount of historical data samples or changing the optimization algorithm. Furthermore, existing technologies, in the optimization process, place four types of construction parameters—jetting pressure, slurry flow rate, drill pipe lifting speed, and drill pipe rotation speed—in the same dimension for simultaneous optimization, without distinguishing the different physical roles of each parameter in actual construction control. Among them, jetting pressure and slurry flow rate are active control variables that are dynamically adjusted with depth, while drill pipe lifting speed and rotation speed are conditional variables constrained by construction efficiency. Undifferentiated overall optimization of these two types of parameters with different properties leads to a mismatch between the convergence direction of the optimization process and the actual adjustable degrees of freedom of the construction equipment. The resulting parameter combination exhibits a disconnect between theory and practice at the construction implementation level.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide an optimized control method for high-pressure jet grouting pile construction to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for optimizing and controlling the construction of high-pressure jet grouting piles, comprising the following steps: S1: Divide the target pile location into several continuous depth layers, obtain the soil type code and permeability coefficient of each depth layer, and determine the target depth layer to be optimized. S2: Use the slurry flow rate and injection pressure of the adjacent depth layer above the target depth layer as the initial value of the current layer boundary. If the current depth layer is the first layer, then use the preset initial slurry flow rate and initial injection pressure as the initial value of the boundary. S3: Combining the soil type, permeability coefficient, initial boundary value, and initial construction parameter vector of the target depth layer, construct a grout diffusion model based on fluid-structure interaction theory; S4: Use the grout diffusion model to solve the dynamic pressure field of the grout inside the soil at the target depth layer, generate the spatial distribution matrix of grout saturation based on the dynamic pressure field, input the spatial distribution matrix into the pre-calibrated strength mapping relationship model, and perform weighted calculations in combination with the soil category code of the target depth layer to obtain the predicted value of the pile compressive strength. S5: Compare the predicted value of the pile compressive strength with the preset target strength value. If the deviation exceeds the limit, activate the gradient descent correction mechanism to dynamically correct the initial value of the boundary of the current depth layer. S6: Repeat steps S2 to S5 to update the initial boundary values ​​of the slurry diffusion model until the deviation converges, and obtain the optimal construction parameters of the current depth layer. Only the slurry flow rate and injection pressure are extracted as the initial boundary values ​​for the next adjacent depth layer iteration calculation. S7: Switch the target depth layer layer by layer along the depth direction of the pile body, repeat steps S2 to S6 until all depth layers are traversed, and generate the optimal construction parameter sequence for the whole pile according to the depth order of each depth layer.

[0007] Furthermore, the soil type codes and permeability coefficients for each depth layer are obtained, and the specific steps are as follows: The target pile location is divided into several continuous depth layers at equal intervals according to a preset fixed layer thickness along the depth direction of the pile location. In-situ static cone penetration test and standard penetration test were conducted on the geological exploration borehole at the target pile location to obtain cone tip resistance, side friction resistance and standard penetration blow count; For each depth layer, the soil type code is determined based on the cone tip resistance, side friction resistance, and standard penetration test blow count, and the permeability coefficient of the depth layer is calibrated by field borehole water injection test or variable head permeability test.

[0008] Furthermore, the method for determining the soil category code is as follows: A ratio-type parameter is constructed to characterize the particle characteristics of soil. The ratio-type parameter is determined by the cone tip resistance and the side friction resistance. A set of boundary values ​​for soil category classification is predefined. The set of boundary values ​​includes at least the upper and lower threshold values ​​for ratio parameters, as well as the upper and lower threshold values ​​for standard penetration test blow counts. Based on the comparison results of the ratio class parameters and the standard penetration test blow count relative to the limit value, the preset soil category code corresponding to the soil at this depth layer is determined by matching; If the conditions for determining any of the preset soil category codes are not met, the soil at that depth layer will be classified as the preset catch-all soil category code.

[0009] Furthermore, the slurry diffusion model based on fluid-structure interaction theory is constructed and used to output the spatial distribution matrix of slurry saturation at the target depth layer in the following manner, with specific steps as follows: The computational domain of the target depth layer is discretized using the finite volume method, resulting in a discrete grid composed of multiple grid nodes, with each grid node corresponding to a spatial location within the computational domain. The injection pressure in the initial value of the target depth layer boundary is used as the slurry pressure boundary condition of the grid node where the nozzle is located. The drill rod lifting speed and rotation speed in the initial construction parameter vector are used to set the nozzle depth movement rate and circumferential rotation rate, respectively. The absolute permeability of the soil is obtained by converting the permeability coefficient of the target depth layer, and the relative permeability of the grout corresponding to the target depth layer is determined from the pre-established correspondence between the soil category code and the relative permeability of the grout. Establish the slurry phase continuity equation, slurry phase seepage equation, and soil skeleton deformation equation to form a slurry diffusion model based on fluid-structure interaction theory; Solve the fluid-structure interaction control equations simultaneously to obtain the grout pressure and soil displacement at each grid node, and calculate the grout saturation at each grid node based on the grout pressure. When the changes in grout pressure and soil displacement of all grid nodes are less than the corresponding preset convergence thresholds, the calculation is determined to have reached a steady state. The grout saturation of each grid node under steady state is arranged according to the spatial position of the grid node to form a spatial distribution matrix of grout saturation.

[0010] Furthermore, the spatial distribution matrix of grout saturation is input into a pre-calibrated strength mapping model, and weighted calculations are performed in conjunction with the soil category code of the target depth layer to obtain the predicted value of the pile compressive strength. The specific steps are as follows: Collect multiple sets of historical high-pressure jet grouting pile engineering data, and obtain the measured values ​​of pile core compressive strength, grout saturation value and soil type code for each grid node at different depth layers; Using the slurry saturation value and soil type code at each grid node as input, and the measured core compressive strength of the pile as output, a strength mapping relationship model under different soil type codes is constructed. Extract the slurry saturation of each grid node in the spatial distribution matrix of slurry saturation, and combine it with the soil category code of the target depth layer to input the strength mapping relationship model. Define the pile compressive strength calculation result output by the strength mapping relationship model for each grid node as the intermediate value of pile compressive strength. Based on the preset weight coefficients of the soil category codes corresponding to each grid node, the median value of the pile compressive strength of each grid node is weighted to obtain the weighted value of the pile compressive strength of each grid node. The weighted average value of the pile compressive strength of all grid nodes within the target depth layer is calculated to obtain the predicted value of the pile compressive strength of the target depth layer.

[0011] Furthermore, the intensity mapping relationship model adopts different function structures for encoding different soil types; the function structures are selected from a pre-established set of function structures, which includes at least exponential functions, linear functions, and piecewise functions. The fitting parameters in each function were determined by performing regression analysis on historical data belonging to the same soil category code.

[0012] Furthermore, when the deviation exceeds the limit, the gradient descent correction mechanism is activated, and the specific steps are as follows: The squared deviation between the predicted pile compressive strength value at the target depth layer and the preset target strength value is taken as the loss value; Calculate the partial derivative of the loss value with respect to the injection pressure in the initial boundary value to obtain the injection pressure gradient; calculate the partial derivative of the loss value with respect to the slurry flow rate in the initial boundary value to obtain the slurry flow rate gradient. Based on the soil type code of the target depth layer, the corresponding injection pressure response coefficient and grout flow rate response coefficient are determined by looking up a table from the pre-established correspondence between soil type codes and response coefficients. The injection pressure gradient is multiplied by the injection pressure response coefficient, and the product is defined as the injection pressure update step size. The slurry flow rate gradient is multiplied by the slurry flow rate response coefficient, and the product is defined as the slurry flow rate update step size. Subtract the injection pressure update step from the current injection pressure, and subtract the slurry flow update step from the current slurry flow rate to complete the iterative update of the boundary initial values. During the iterative update process, the drill rod lifting speed and drill rod rotation speed in the boundary initial values ​​always remain constant at the original values ​​of the preset initial construction parameter vector.

[0013] Furthermore, the initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, thus obtaining the optimal construction parameters for the current depth layer. The specific steps are as follows: Based on the iteratively updated initial boundary values, the grout diffusion model is reconstructed. The dynamic pressure field and the spatial distribution matrix of grout saturation are calculated using the reconstructed grout diffusion model. The spatial distribution matrix of grout saturation is then processed using a strength mapping relationship model to obtain new predicted values ​​of pile compressive strength. Calculate the deviation between the new predicted value of pile compressive strength and the preset target strength value; stop the iteration when the absolute value of the deviation is less than or equal to the preset convergence threshold; determine the injection pressure and grout flow rate at the time of stopping the iteration as the optimal injection pressure and optimal grout flow rate for the target depth layer; The drill rod lifting speed in the preset initial construction parameter vector is determined as the optimal drill rod lifting speed for the target depth layer; the drill rod rotation speed in the preset initial construction parameter vector is determined as the optimal drill rod rotation speed for the target depth layer; The optimal construction parameters for the target depth layer consist of the optimal injection pressure, optimal slurry flow rate, optimal drill pipe lifting speed, and optimal drill pipe rotation speed.

[0014] Furthermore, the target depth layer is switched layer by layer along the depth direction of the pile until all depth layers of the entire pile are traversed. The specific steps are as follows: The first layer is taken as the current depth layer. The preset initial slurry flow rate and preset initial injection pressure are used as the initial boundary values ​​of the current depth layer. The gradient descent correction mechanism is executed and the initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, so as to obtain the optimal construction parameters of the first layer. Extract the optimal slurry flow rate and optimal injection pressure of the current depth layer, and use them as the initial values ​​of the iteration boundary of the next depth layer; The layers are iterated according to the depth arrangement order, the next depth layer is determined as the new current depth layer, the gradient descent correction mechanism is repeatedly executed and the initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, and the optimal construction parameters of the depth layer are obtained. The process is iterated step by step until the parameters of the last layer at the bottom of the pile are optimized. The optimal construction parameters for all depth layers are arranged in order of depth to generate a sequence of optimal construction parameters for the entire pile.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention utilizes fluid-structure interaction theory to construct a grout diffusion model during the layer-by-layer optimization process. It incorporates the soil type, permeability coefficient, initial boundary values, and construction parameter vectors of the target depth layer into a unified physical calculation framework. This allows the dynamic pressure field and saturation spatial distribution of the grout in the soil to be accurately solved, providing input conditions that conform to the actual physical process for subsequent strength prediction. This solves the problem that the pure data-driven model in the prior art is not responsive enough to changes in formation conditions. This invention introduces a unit-by-unit weighted calculation of the spatial distribution matrix of grout saturation and soil characteristics into the strength prediction process. This incorporates the difference in strength contribution between the actual filling state of the grout at various spatial locations and the corresponding soil unit, allowing the predicted pile compressive strength to reflect the impact of the non-uniformity of grout distribution on the overall strength, thus improving the accuracy of strength prediction. When the deviation exceeds the limit, a gradient descent correction mechanism is activated, and the gradient correction amplitude is adaptively adjusted according to the response coefficient corresponding to the soil type. This ensures that the direction and step size of parameter correction match the current stratum conditions, avoiding the slow convergence or over-adjustment problems that occur in different strata when the fixed step size correction is applied. This invention uses only the optimal grout flow rate and optimal injection pressure after the current depth layer converges as the constituent parameters of the initial value of the boundary of the next depth layer. This allows the actual grout injection state at the end of the construction of the previous depth layer to directly participate in the initial condition setting of the next depth layer. From a physical mechanism perspective, this ensures the interlayer continuity and executability of the construction parameter sequence for the entire pile length, and effectively avoids abrupt changes in pile strength along the depth direction or the formation of weak interlayers caused by jumps in interlayer parameters. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a curve showing the strength deviation versus loss value fitting for the present invention. Figure 3 This is a scatter plot of the spatial penalty term and loss value of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0018] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] Example: Please see Figures 1-3 The present invention provides a technical solution: A method for optimizing and controlling the construction of high-pressure jet grouting piles, comprising the following steps: S1: Divide the target pile location into several continuous depth layers, obtain the soil type and permeability coefficient of each depth layer, and determine the target depth layer to be optimized. In a specific embodiment of the present invention, considering that the construction of high-pressure jet grouting piles is a dynamic process that is continuously advanced along the depth direction, the physical and mechanical properties and permeability characteristics of the soil at different depths are different. These differences will directly affect the diffusion range of the grout in the soil and the final pile strength. After dividing the depth layers along the pile depth direction, the stratum conditions of each depth layer can be independently modeled and the parameters optimized, so that the setting of construction parameters can be accurately matched with the changes in stratum, avoiding the local strength deficiency or grout waste caused by using uniform parameters for the whole pile. The specific steps to obtain the soil category code and permeability coefficient for each depth layer are as follows: The target pile location is divided into several continuous depth layers at equal intervals according to a preset fixed layer thickness along the depth direction of the pile location. The aforementioned depth layer refers to a series of continuously arranged horizontal layered units that divide the depth range of the pile body from top to bottom according to a preset layer thickness along the depth direction of the target pile location. Each depth layer has its own top and bottom elevations, and adjacent depth layers are connected end to end. The preset layer thickness is a single fixed value determined based on the scale of strata changes revealed by geological exploration boreholes. The principle for selecting this fixed value is to keep the soil type and permeability coefficient within a single depth layer uniform. If the layer thickness is selected too large, multiple soil types will be included in the same depth layer, leading to a decrease in the accuracy of subsequent grout diffusion calculations and strength predictions. If the layer thickness is selected too small, the number of depth layers for the entire pile will be too large, reducing calculation efficiency. For each depth layer, the soil type and permeability coefficient of that depth layer need to be obtained through in-situ tests. Soil type refers to the classification of soil based on its particle composition and physical state. In this invention, soil types include at least four categories: silty clay, silty clay, sandy soil, and gravelly soil. Permeability coefficient is a parameter that indicates the ability of soil to allow groundwater or grout to pass through its pores. The magnitude of the permeability coefficient directly determines the diffusion rate and diffusion range of grout in the soil and is a key input parameter for the grout diffusion model. The specific procedures for obtaining soil type and permeability coefficient are as follows: In-situ static cone penetration test and standard penetration test are conducted on the geological exploration borehole at the target pile location. The static cone penetration test involves pressing a conical probe into the soil at a constant rate and measuring the cone tip resistance and side friction resistance experienced by the probe. The standard penetration test involves hammering a standard penetrator into the soil from the bottom of the borehole and recording the number of hammer blows required to penetrate to a certain depth as the standard penetration blow count. Both of the above tests are routine in-situ testing methods in the field of geotechnical engineering investigation, and their testing equipment and operating procedures should comply with the provisions of the current relevant national technical standards. For each depth layer, the soil type code of that depth layer is determined based on cone tip resistance, side friction resistance and standard penetration test blow count, and the permeability coefficient of that depth layer is calibrated by field borehole water injection test or variable head permeability test. For each depth layer, the cone tip resistance and side friction resistance within that depth layer are extracted from the static cone penetration test data. The cone tip resistance is divided by the side friction resistance, and the resulting ratio is defined as the friction ratio. The friction ratio is a dimensionless parameter used to characterize the fineness of soil particles. Its physical meaning is: a lower friction ratio indicates a higher clay content and finer soil particles; a higher friction ratio indicates a higher sand content and coarser soil particles. The standard penetration test (SPT) blow counts within this depth range are extracted from the SPT data. The SPT blow count is a parameter reflecting the compactness and strength characteristics of the soil. A lower SPT blow count indicates that the soil is relatively weak, while a higher SPT blow count indicates that the soil is relatively dense and hard. The friction ratio and SPT blow count are compared with a pre-defined set of soil category classification thresholds. This set of thresholds was determined through statistical analysis of a large amount of historical geological survey data. Specifically, it includes the upper limit thresholds for friction ratio and SPT blow count for silty clay, the lower limit thresholds for friction ratio and upper limit thresholds for friction ratio and lower limit thresholds for SPT blow count for silty clay, and the lower limit thresholds for friction ratio and lower limit thresholds for SPT blow count for sandy soil. The upper limit threshold refers to the maximum value allowed for the corresponding parameter when determining a certain type of soil, and the lower limit threshold refers to the minimum value allowed for the corresponding parameter when determining a certain type of soil. When the friction ratio is less than the upper limit threshold of the friction ratio corresponding to silty clay and the SPT blow count is less than the upper limit threshold of the SPT blow count corresponding to silty clay, it indicates that the soil at this depth has the characteristics of low friction ratio and low SPT blow count, that is, the soil particles are fine and the strength is low, which is consistent with the engineering characteristics of silty clay. The soil type at this depth is determined to be silty clay. When the friction ratio is not less than the lower limit threshold of the friction ratio corresponding to silty clay and not greater than the upper limit threshold of the friction ratio corresponding to silty clay, and the SPT blow count is not less than the lower limit threshold of the SPT blow count corresponding to silty clay and not greater than the upper limit threshold of the SPT blow count corresponding to silty clay, it indicates that the soil at this depth has the characteristics of medium friction ratio and medium SPT blow count, which is consistent with the engineering characteristics of silty clay. The soil type at this depth is determined to be silty clay. When the friction ratio is greater than the lower limit threshold of the friction ratio corresponding to sandy soil and the SPT blow count is greater than the lower limit threshold of the SPT blow count corresponding to sandy soil, it indicates that the soil at this depth has the characteristics of high friction ratio and high SPT blow count, that is, the soil particles are coarse and relatively dense, which is consistent with the engineering characteristics of sandy soil. The soil type at this depth is determined to be sandy soil. When the friction ratio and SPT blow count of a certain depth layer do not meet the criteria for determining any of the soil types, such as silty clay, fine clay, or sandy soil, it indicates that the soil at that depth layer does not belong to the above three conventional soil types. In this case, the soil type of the depth layer is determined to be gravelly soil. Gravelly soil is a catch-all category that covers a type of soil with larger particle size and lack of cohesion between particles. Its engineering characteristics are significantly different from the first three types of soil. A special treatment method will be used in the subsequent slurry diffusion calculation and strength prediction. To facilitate the retrieval and processing of soil type information in subsequent steps, a unique soil category code is assigned to each of the above four soil types. The soil category code can be a numerical number or a character identifier, as long as the codes corresponding to different soil types are different. The introduction of soil category codes allows the soil type, a classification variable, to participate in subsequent model calculations in numerical form, simplifying the data processing flow. The soil category code is one of several preset category codes, which include at least a first category code, a second category code, a third category code, and a catch-all category code. The soil category code is determined by static cone penetration test data and standard penetration test data according to preset threshold judgment rules. Specifically, the method for determining soil category codes is as follows: A ratio-type parameter is constructed to characterize the particle characteristics of soil. The ratio-type parameter is determined by the cone tip resistance and the side friction resistance. A set of boundary values ​​for soil category classification is predefined. The set of boundary values ​​includes at least the upper and lower threshold values ​​for ratio parameters, as well as the upper and lower threshold values ​​for standard penetration test blow counts. Based on the comparison results of the ratio class parameters and the standard penetration test blow count relative to the limit value, the preset soil category code corresponding to the soil at this depth layer is determined by matching; If the determination criteria of any preset soil category code are not met, the soil at this depth layer shall be classified as the preset catch-all soil category code. After determining the soil type, the permeability coefficient of the soil at each depth layer is determined by either a field borehole water injection test or a variable head permeability test. The field borehole water injection test involves injecting a certain amount of water into the borehole and observing the injection rate and water level changes within the borehole. The permeability coefficient is then calculated based on the stable relationship between the injection rate and water level changes. The variable head permeability test involves loading undisturbed soil samples collected in the field into an indoor permeameter and measuring the rate of water level drop in the variable head pipe over a certain period of time. The permeability coefficient is then calculated by combining the soil sample cross-sectional dimensions and the head difference. The choice between the two test methods should be determined based on the field conditions and soil type to ensure that the accuracy of the permeability coefficient measurement meets the requirements of subsequent slurry diffusion calculations. Soil category codes and permeability coefficients for each depth layer are arranged sequentially from shallowest to deepest, forming soil category code sequences and permeability coefficient sequences respectively. The a-th element in these two sequences corresponds to the soil category code and permeability coefficient of the a-th depth layer, thereby establishing a profile of ground property parameters in the direction of pile depth. After completing the above steps, starting from the first depth layer, it is determined as the target depth layer to be optimized. The target depth layer refers to the depth layer where grout diffusion calculation, strength prediction and construction parameter optimization are currently being performed. Each optimization iteration is only performed on one target depth layer. After optimization is completed, the target depth layer is switched to the next depth layer until all depth layers of the entire pile have been optimized. It should be noted that the target depth layer is a dynamic concept that advances layer by layer along the depth direction. When the target depth layer is the first depth layer, its initial boundary values ​​adopt the preset initial grout flow rate and preset initial injection pressure. When the target depth layer is not the first layer, the grout flow rate and injection pressure in its initial boundary values ​​adopt the optimal grout flow rate and optimal injection pressure obtained when the previous depth layer was optimized. This interlayer transfer mechanism allows the actual grout injection state at the end of the construction of the previous depth layer to directly participate in the initial condition setting of the next depth layer. From a physical mechanism perspective, this ensures the interlayer continuity of the entire pile construction parameter sequence and avoids abrupt changes in pile strength along the depth direction or the formation of weak interlayers caused by interlayer parameter jumps.

[0020] S2: Use the slurry flow rate and injection pressure of the adjacent depth layer above the target depth layer as the initial value of the current layer boundary. If the current depth layer is the first layer, then use the preset initial slurry flow rate and initial injection pressure as the initial value of the boundary. In this step, it is necessary to determine the initial boundary values ​​for the target depth layer to be optimized. The initial boundary values ​​refer to the initial injection pressure value and the initial slurry flow rate value input into the slurry diffusion model to start the slurry diffusion calculation. These two values ​​together constitute the pressure boundary condition and flow rate boundary condition of the slurry diffusion model at the nozzle. The above-mentioned initial value of the current layer boundary is the initial value of the current target depth layer boundary. The determination of the initial value of the current layer boundary follows the principle of inter-layer state transfer, specifically: determining whether the current target depth layer is the first depth layer. The first depth layer refers to the depth layer located at the shallowest position and corresponding to the starting depth range of the pile top among several consecutive depth layers divided along the pile depth direction. If the current target depth layer is the first depth layer and there are no adjacent depth layers that have been optimized above it, the injection pressure in the initial value of the current layer boundary is taken as the preset initial injection pressure, and the slurry flow rate in the initial value of the current layer boundary is taken as the preset initial slurry flow rate. The preset initial injection pressure and the preset initial slurry flow rate are construction start parameters determined in the construction preparation stage based on engineering experience, equipment rated working parameters and soil type of the stratum. Their specific values ​​are preset in the construction plan. If the current target depth layer is not the first depth layer, there is an adjacent depth layer above it that has completed slurry diffusion calculation and parameter optimization. This adjacent depth layer is defined as the previous adjacent depth layer. After optimization convergence, the optimal injection pressure and optimal slurry flow rate of the previous adjacent depth layer have been determined. The optimal injection pressure of the previous adjacent depth layer is used as the injection pressure in the initial value of the current layer boundary, and the optimal slurry flow rate of the previous adjacent depth layer is used as the slurry flow rate in the initial value of the current layer boundary. This processing method allows the actual slurry injection state at the end of the construction of the previous depth layer to be directly transferred to the initial calculation conditions of the next depth layer, thus establishing physical continuity between adjacent depth layers. The above-mentioned grout flow rate refers to the volume of solidified grout injected into the soil through the nozzle of the high-pressure jet grouting drill rod per unit time, and the injection pressure refers to the pressure value output by the solidified grout at the nozzle. Both are active control parameters that are adjusted in real time by the operator or the automatic control system during the construction of high-pressure jet grouting piles.

[0021] S3: Combining the soil type, permeability coefficient, initial boundary value, and initial construction parameter vector of the target depth layer, construct a grout diffusion model based on fluid-structure interaction theory; The initial construction parameter vector is a two-dimensional parameter set consisting of a preset drill rod lifting speed and a preset drill rod rotation speed. The drill rod lifting speed refers to the linear velocity of the drill rod as it is lifted upward along the depth direction during high-pressure jet grouting. The drill rod rotation speed refers to the angular velocity of the drill rod as it rotates around its own axis during high-pressure jet grouting. Both the drill rod lifting speed and the drill rod rotation speed are construction parameters that are preset in the construction preparation stage and remain constant during the optimization process of a single depth layer. The specific values ​​of the preset drill rod lifting speed and the preset drill rod rotation speed are determined comprehensively based on the rated working capacity of the construction equipment, the pile diameter design requirements, and the construction efficiency requirements. The slurry diffusion model based on fluid-structure interaction theory is constructed and used to output the spatial distribution matrix of slurry saturation at the target depth layer in the following manner: The grout diffusion model based on fluid-structure interaction theory is a mathematical model that simultaneously considers the coupling effect between the flow behavior of grout in porous media and the deformation behavior of the soil skeleton under grout pressure. Fluid-structure interaction refers to the two-way interaction between fluid motion and solid deformation; that is, changes in grout pressure cause changes in the effective stress of the soil skeleton, leading to displacement and changes in pore volume. These changes in pore volume, in turn, affect the seepage channels and pressure distribution of the grout within the pores. By introducing the fluid-structure interaction mechanism, the cutting, displacement, and squeezing effects of grout on the soil during high-pressure jet grouting can be more realistically reflected, thus obtaining a more accurate spatial distribution of grout saturation. The slurry diffusion model based on fluid-structure interaction theory is constructed through the following steps: The computational domain of the target depth layer is discretized using the finite volume method, resulting in a discrete grid composed of multiple grid nodes, with each grid node corresponding to a spatial location within the computational domain. The computational domain refers to the spatial range corresponding to the target depth layer. This spatial range is a cylindrical area formed by expanding radially outward by a preset distance based on the design pile diameter of the high-pressure jet grouting pile. The principle for selecting the expansion distance is to ensure that the grout pressure gradient at the boundary of the computational domain has been attenuated to a negligible level. The finite volume method is a numerical calculation method that integrates continuous partial differential equations over discrete control volumes. Its basic idea is to divide the computational domain into several non-overlapping control volumes. Each control volume revolves around a grid node. By integrating the control equations over each control volume, the differential equations are transformed into a system of algebraic equations for solution. After discretization using the finite volume method, a discrete grid consisting of multiple grid nodes is obtained. The discrete grid refers to the spatial grid structure formed after the computational domain is discretized. The grid node refers to the center point or vertex of each control volume in the discrete grid. Each grid node corresponds to a spatial location in the computational domain. This spatial location is determined by the depth coordinates and the two-dimensional coordinates in the horizontal plane. The soil property parameters and the physical quantities to be solved are stored at each grid node. The injection pressure in the initial value of the target depth layer boundary is used as the slurry pressure boundary condition of the grid node where the nozzle is located. The drill rod lifting speed and rotation speed in the initial construction parameter vector are used to set the nozzle depth movement rate and circumferential rotation rate, respectively. The nozzle refers to the outlet position of the solidified grout injected onto the high-pressure rotary grouting drill rod. The grid node where the nozzle is located refers to the grid node in the discrete grid that coincides with or is closest to the spatial position of the nozzle. The grout pressure boundary condition refers to the grout pressure value that is forcibly specified on the grid node where the nozzle is located during the numerical solution of the grout diffusion model. This value remains constant throughout the solution process and is equal to the injection pressure in the initial value of the current layer boundary. The nozzle depth-direction movement rate refers to the linear velocity of the nozzle moving along the depth direction. Its value is equal to the drill rod lifting speed in the initial construction parameter vector. During the numerical solution of the grout diffusion model, the nozzle depth-direction movement rate is used to determine the depth position of the nozzle at different times. The circumferential rotation rate of the nozzle refers to the angular velocity of the nozzle rotating around the drill rod axis. Its value is equal to the drill rod rotation speed in the initial construction parameter vector. In the numerical solution of the slurry diffusion model, the circumferential rotation rate of the nozzle is used to determine the horizontal direction pointed by the nozzle at different times. The absolute permeability of the soil is obtained by converting the permeability coefficient of the target depth layer, and the relative permeability of the grout corresponding to the target depth layer is determined from the pre-established correspondence between the soil category code and the relative permeability of the grout. Absolute soil permeability refers to the soil's ability to allow a single fluid to pass through it. Its value is only related to the geometric characteristics of the soil's pore structure and is independent of the properties of the fluid passing through it. The conversion relationship between permeability coefficient and absolute soil permeability is as follows: multiply the permeability coefficient by the dynamic viscosity of the fluid passing through it, and then divide by the product of gravitational acceleration and fluid density to obtain the absolute soil permeability. In this step, the fluid passing through it is groundwater, and its dynamic viscosity, density, and gravitational acceleration are all based on standard values ​​under normal temperature conditions. The relative permeability of grout refers to the reduction factor of the effective seepage capacity of grout in soil relative to the absolute permeability of the soil when two-phase fluids of grout and groundwater exist in the pores of the soil. This factor is a dimensionless value between 0 and 1. The value of the relative permeability of grout is related to the soil type. Different soil types have different pore structures and particle surface properties, and their resistance to grout also varies. The pre-established correspondence between soil type codes and relative permeability of grout is formed by conducting indoor grout seepage tests on different soil types, measuring the relative permeability of grout under different grout saturation conditions, and compiling the data into a lookup table or function relationship. During model calculation, the corresponding relative permeability value of grout is directly called according to the soil type code. Establish the slurry phase continuity equation, slurry phase seepage equation, and soil skeleton deformation equation to form a slurry diffusion model based on fluid-structure interaction theory; The slurry phase continuity equation is a partial differential equation describing the conservation of slurry mass. Its physical meaning is: within any control volume, the rate of change of slurry mass with time is equal to the net mass of slurry flowing into or out of the control volume boundary per unit time. The slurry phase continuity equation contains the spatiotemporal partial differential relationship between three physical quantities: slurry density, slurry saturation, and slurry Darcy velocity. Slurry density refers to the mass of solidified slurry per unit volume, slurry saturation refers to the proportion of the volume occupied by slurry in the soil pore space to the total pore volume, and slurry Darcy velocity refers to the volumetric flow rate of slurry per unit cross-sectional area in the soil pores. The grout phase flow equation describes the motion of grout in porous media. Specifically, it is a functional relationship between the Darcy velocity of the grout and the absolute permeability of the soil, the relative permeability of the grout, the dynamic viscosity of the grout, the pressure gradient of the grout, and the gravity vector. The dynamic viscosity of the grout refers to the physical quantity of the internal frictional resistance of the solidified grout during the flow process. Its value is related to the mix ratio and temperature of the solidified grout. The pressure gradient of the grout refers to the rate of change of the grout pressure in various directions in space. The spatial distribution difference of the grout pressure is the fundamental reason driving the flow of the grout. The gravity vector is a vector that is vertically downward and has the magnitude of the gravitational acceleration. It is used to characterize the influence of gravity on the flow of the grout. The soil skeleton deformation equation is a mechanical equation that describes the deformation of soil under the combined action of grout pressure and external force. Its physical meaning is that the change in stress state of the soil skeleton is equal to the change in effective stress, and the change in effective stress is affected by the change in grout pressure. The displacement distribution of soil under grout pressure can be calculated through the soil skeleton deformation equation. Soil displacement refers to the amount of movement of each spatial position in the soil skeleton relative to the initial position. The grout phase continuity equation, grout phase seepage equation, and soil skeleton deformation equation are combined to form a fluid-structure interaction control equation set. The solution process of this control equation set is as follows: based on the grout pressure distribution and soil displacement distribution at the current moment, the grout phase continuity equation and seepage equation are solved simultaneously to obtain the grout pressure distribution and grout saturation distribution at the next moment. Then, based on the change in grout pressure, the change in soil displacement is calculated by substituting it into the soil skeleton deformation equation, and the pore structure parameters of the soil are updated. The updated parameters are then fed back into the seepage calculation in the next solution step. This process is repeated alternately until the grout pressure field and soil displacement field no longer change significantly, that is, a steady state is reached.

[0022] S4: Use the grout diffusion model to solve the dynamic pressure field of the grout inside the soil at the target depth layer, generate the spatial distribution matrix of grout saturation based on the dynamic pressure field, input the spatial distribution matrix into the pre-calibrated strength mapping relationship model, and perform weighted calculations in combination with the soil category code of the target depth layer to obtain the predicted value of the pile compressive strength. The dynamic pressure field refers to the spatial distribution of grout pressure values ​​at each grid node in the computational domain of the target depth layer during the solution process of the grout diffusion model. Since high-pressure jetting is a transient process, after the grout is ejected from the nozzle, it diffuses into the pores of the surrounding soil. The grout pressure gradually decreases from the nozzle position to the surrounding area until it reaches a steady state. The spatial distribution of grout pressure during this process is the dynamic pressure field. Solve the fluid-structure interaction control equations simultaneously to obtain the grout pressure and soil displacement at each grid node, and calculate the grout saturation at each grid node based on the grout pressure. Simultaneous solution of the fluid-structure interaction control equations refers to solving the slurry phase continuity equation, slurry phase seepage equation, and soil skeleton deformation equation as a whole in each solution step, rather than solving them independently and then exchanging the data. Simultaneous solution ensures that the coupling relationship between the various physical fields is satisfied in each solution step, thereby improving the calculation accuracy. By simultaneously solving the fluid-structure interaction control equations, the slurry pressure value and soil displacement value at each grid node on the discrete grid are obtained. The slurry pressure value refers to the pressure exerted by the slurry per unit area at that grid node, and the soil displacement value refers to the spatial movement of the soil skeleton at that grid node from its initial position. Grout saturation refers to the ratio of the volume of grout in the pore space corresponding to a grid node to the total pore volume. Its value is between 0 and 1. The grout saturation is calculated by substituting the grout pressure value of each grid node at the current moment into the corresponding relationship of the pre-calibrated soil-water characteristic curve. This soil-water characteristic curve describes the nonlinear functional relationship between grout pressure and grout saturation and is determined by indoor tests. When the changes in grout pressure and soil displacement of all grid nodes are less than the corresponding preset convergence thresholds, the calculation is determined to have reached a steady state. The grout saturation of each grid node under steady state is arranged according to the spatial position of the grid node to form a spatial distribution matrix of grout saturation. The change in grout pressure refers to the absolute difference in grout pressure values ​​at the same grid node between two adjacent solutions. The change in soil displacement refers to the magnitude of the vector difference in soil displacement values ​​at the same grid node between two adjacent solutions. The preset convergence thresholds are two pre-set positive real numbers used to determine whether the grout pressure field and the soil displacement field have stopped changing substantially over time. When the change in grout pressure at all grid nodes on the discrete grid is less than the preset convergence threshold corresponding to grout pressure, and the change in soil displacement at all grid nodes is less than the preset convergence threshold corresponding to soil displacement, it is considered that the diffusion process of grout in the soil has become stable, and subsequent solutions will not substantially change the results. At this point, the solution is stopped, and the calculation is considered to have reached a steady state. The steady state refers to the calculation state in which the spatial distribution of grout pressure and soil displacement no longer changes significantly over time. The slurry saturation spatial distribution matrix is ​​a two-dimensional data structure with the horizontal row index and horizontal column index of the grid nodes as two dimensions. Each matrix element stores the slurry saturation value corresponding to the spatial location. The element in the m-th row and n-th column of the slurry saturation spatial distribution matrix corresponds to the slurry saturation value of the grid node in the m-th row and n-th column of the horizontal section where the target depth layer is located in the depth direction in the discrete grid. This slurry saturation spatial distribution matrix can completely describe the degree of filling of slurry at each spatial location in the horizontal section within the target depth layer. After the spatial distribution matrix of grout saturation is generated, it is input into a pre-calibrated strength mapping model. The strength mapping model is a mathematical mapping relationship with the grout saturation value and soil type code at the grid node as input variables and the pile compressive strength at that grid node as the output variable. Its function is to calculate the compressive strength that the cement-soil reinforced body can achieve at that spatial location based on the degree of grout filling and the soil type. The pre-calibration process of the strength mapping model is as follows: Collect multiple sets of historical high-pressure jet grouting pile engineering data, and obtain the measured values ​​of pile core compressive strength, grout saturation value and soil type code for each grid node at different depth layers; Historical high-pressure jet grouting pile project data refers to the pile quality inspection data obtained through field tests and laboratory tests in completed high-pressure jet grouting pile projects. For each set of historical data, it is necessary to obtain the measured core compressive strength, grout saturation value, and soil category code of the pile body at each grid node within the same depth layer. The measured core compressive strength value refers to the compressive strength value obtained after drilling core samples from the completed high-pressure jet grouting pile body and conducting uniaxial compressive strength tests on the core samples according to standard test methods. This compressive strength value reflects the actual mechanical properties of the cement-soil mixture at that spatial location. The grout saturation value refers to the degree of grout filling obtained at the grid node through back calculation using the grout diffusion model or through core sample porosity analysis. The soil category code refers to the code corresponding to the soil type at the depth layer where the grid node is located, and its determination method is the same as that used in step S1. Using the slurry saturation value and soil category code at each grid node as input, and the measured core compressive strength value of the pile corresponding to that grid node as output, a strength mapping relationship model under different soil category codes is constructed. The strength mapping relationship model refers to the functional correspondence between input variables and output variables established by regression analysis. Its essence is a set of one or more mathematical functions. Given the specific values ​​of the input variables, the corresponding output value can be calculated through this set of functions. Since the mixing and consolidation mechanism and strength development law of cement slurry and soil particles differ under different soil types, it is necessary to establish a corresponding strength mapping relationship model for each soil category code. The strength mapping relationship model under each soil category code adopts a different function structure. The function structure refers to the basic mathematical function form used in the regression model. When the soil category code corresponds to silty clay or silty clay, the strength mapping relationship model adopts an exponential function structure, that is, the compressive strength of the pile increases exponentially with the increase of grout saturation. This is because the chemical reaction and cementation between grout and soil particles in cohesive soil accelerates the strength growth rate after reaching a certain grout saturation. When the soil category code corresponds to sandy soil, the strength mapping relationship model adopts a linear function structure, that is, the compressive strength of the pile increases linearly with the increase of grout saturation. This is because the grout in sandy soil mainly fills the pores and cements the particles, and the strength growth is basically proportional to the degree of grout filling. When the soil category code corresponds to gravelly soil, the strength mapping relationship model adopts a piecewise function structure to adapt to the inconsistent strength growth law of gravelly soil in different slurry saturation ranges. The fitting parameters in each function structure, including the base coefficient and exponential coefficient of the exponential function, the slope and intercept of the linear function, and the boundary values ​​and coefficients within each segment of the piecewise function, are all determined by least squares regression analysis on all historical data belonging to the same soil category code. After the strength mapping model is calibrated, the application phase begins. The slurry saturation values ​​of each grid node in the slurry saturation spatial distribution matrix are extracted. The slurry saturation value of each grid node, along with the soil category code of the target depth layer where that grid node is located, is input into the strength mapping model corresponding to that soil category code. The strength mapping model then outputs the calculated pile compressive strength for that grid node. This calculated result is defined as the intermediate value of the pile compressive strength. The intermediate value of the pile compressive strength refers to the estimated compressive strength directly output by the strength mapping model based on the slurry saturation value and soil category code of a single grid node. This estimated compressive strength only reflects the contribution of the slurry filling degree and soil type to the strength at that node, and does not yet consider the influence of differences in soil characteristics at different spatial locations within the same depth layer on the overall strength contribution. Based on the preset weighting coefficients of the soil category codes corresponding to each grid node, the median value of the pile compressive strength of each grid node is weighted. The preset weighting coefficients are coefficients pre-set according to different soil category codes to characterize the contribution of a single grid node to the overall pile strength of the layer at its depth under the soil type condition. The determination of the preset weighting coefficients includes factors that affect the uniformity of mixing and consolidation effect of cement grout and soil, such as the compaction degree or clay content of the soil corresponding to the soil category code. The specific weighting operation is as follows: multiply the median value of the pile compressive strength of each grid node by the preset weighting coefficient corresponding to that grid node to obtain the weighted value of the pile compressive strength of each grid node. The weighted value of the pile compressive strength refers to the strength contribution value of the grid node after the weighting coefficient correction. This value comprehensively reflects the combined influence of the grout filling degree, soil type and soil characteristic differences on the strength contribution at the node. The arithmetic mean of the weighted compressive strength values ​​of all grid nodes within the target depth layer is calculated. This involves summing the weighted compressive strength values ​​of all grid nodes within the target depth layer and then dividing by the total number of grid nodes within the target depth layer. The result is the predicted compressive strength value of the pile at the target depth layer. The predicted compressive strength value is a single numerical value representing the overall compressive strength level of the target depth layer, obtained after weighted averaging. This value serves as the basis for determining whether the current construction parameters meet the design requirements and is used for subsequent deviation calculations and parameter corrections.

[0023] S5: Compare the predicted value of the pile compressive strength with the preset target strength value. If the deviation exceeds the limit, activate the gradient descent correction mechanism to dynamically correct the initial value of the boundary of the current depth layer. The preset target strength value refers to the design value of the pile compressive strength determined in advance according to the design drawings and engineering technical requirements of the high-pressure jet grouting pile. This value represents the minimum compressive strength standard that the pile should reach at a certain depth after construction. The preset target strength value is determined by the engineering designers before construction begins based on the load requirements of the superstructure and the foundation treatment design specifications, and remains constant throughout the entire construction optimization and control process. Deviation refers to the difference between the predicted value of the pile compressive strength and the preset target strength value. The positive or negative sign of the deviation indicates the direction of deviation of the predicted strength relative to the target strength. When the predicted value of the pile compressive strength is less than the preset target strength value, the deviation is positive, indicating that the current construction parameters are insufficient to make the pile strength meet the design requirements. When the predicted value of the pile compressive strength is greater than the preset target strength value, the deviation is negative, indicating that the current construction parameters will lead to over-reinforcement and material waste. The absolute value of the deviation indicates the degree of deviation between the predicted strength and the target strength. The preset allowable range refers to the upper limit of the acceptable absolute value of the deviation. The preset allowable range is a positive real number threshold, and its specific value is determined according to the engineering design requirements. It is set according to the percentage of the preset target strength value. The purpose of the preset allowable range is to provide a convergence criterion for the optimization of construction parameters. When the absolute value of the deviation is less than or equal to the preset allowable range, it is considered that the construction parameters corresponding to the current boundary initial value can meet the design requirements and no further adjustment is needed. Deviation comparison refers to the operation of comparing the predicted value of the pile compressive strength with the preset target strength value. When the absolute value of the deviation is less than or equal to the preset allowable range, the deviation is determined to be within the limit, and the initial value of the boundary of the current depth layer is the optimal initial value of the boundary, and the gradient descent correction mechanism is not activated. When the absolute value of the deviation is greater than the preset allowable range, the deviation is determined to be beyond the limit, indicating that the construction parameters corresponding to the current initial value of the boundary cannot meet the design requirements, and the gradient descent correction mechanism needs to be activated to correct the initial value of the boundary. The gradient descent correction mechanism is an iterative optimization method that uses the partial derivative information of the objective function with respect to the optimization variables to guide the update direction and magnitude of the optimization variables. In this invention, the optimization objective of the gradient descent correction mechanism is to minimize the deviation between the predicted value of the pile compressive strength and the preset target strength value. The optimization variables are the injection pressure and grout flow rate in the initial value of the current depth layer boundary. The specific execution steps of the gradient descent correction mechanism are as follows: The deviation between the predicted compressive strength of the pile at the target depth layer and the preset target strength value is squared. The result of the squared deviation is defined as the loss value. The loss value is a non-negative real number used to quantify the severity of the deviation between the construction parameters and the target strength under the current boundary initial value conditions. The purpose of using the squared deviation as the loss value is to make the gradient of the loss value with respect to the boundary initial value mathematically continuous and differentiable, which facilitates the calculation of subsequent partial derivatives. At the same time, the square operation unifies the positive and negative deviations into non-negative values, avoiding the mutual cancellation of positive and negative deviations during the optimization process. Let the predicted compressive strength of the pile at the target depth be _____. The preset target intensity value is The spatial distribution matrix of slurry saturation is Matrix elements This represents the slurry saturation value at the grid node in the i-th row and j-th column of the discrete grid. M and N are both positive integers, representing the number of rows and columns of the discrete grid, respectively. i is the index of the row of the grid node in the discrete grid, and j is the index of the column of the grid node in the discrete grid. Define strength deviation Strength deviation It is a real number, whose positive or negative sign indicates the direction of deviation of the predicted intensity from the target intensity, and its absolute value indicates the degree of deviation; saturation spatial features are extracted from the slurry saturation spatial distribution matrix, and an insufficient saturation identification matrix W is defined, whose elements... Determined according to the following judgment rules: When Less than the preset saturation threshold hour, The value is 1; when Greater than or equal to the preset saturation threshold hour, The value is 0; preset saturation threshold. It is a positive real number between 0 and 1, determined according to the soil type code C of the target depth layer, representing the minimum grout saturation value that can form effective pile strength under the conditions of this soil type. Grid nodes with a value of 1 form a set of low-saturation nodes. The grid nodes in this set are the spatial locations where the grout filling is insufficient and the contribution to the overall pile strength is weak. Define the proportion of low-saturation nodes The number of grid nodes in the low-saturation node set divided by the total number of grid nodes M in the discrete grid multiplied by N, i.e. The proportion of low-saturation nodes, r, is a dimensionless real number between 0 and 1. The closer it is to 1, the more extensive the area of ​​insufficient slurry filling within the target depth layer; Loss value Construct using the following formula: in, This is the loss value used to drive the gradient descent correction mechanism. This is the space penalty coefficient, a positive real number, whose value is preset based on engineering experience or parameter sensitivity analysis. Increasing the spatial saturation uniformity amplifies the loss value, forcing optimization to focus more on the uniformity of slurry filling. The time loss function degenerates into the square of the pure intensity deviation; When two different initial boundary values ​​produce the same strength deviation At that time, the proportion of low-saturation nodes The larger condition corresponds to a larger one. This generates a stronger gradient correction driving force, enabling optimization to simultaneously reduce strength deviation and improve slurry filling uniformity. Set space penalty coefficient Use the default value of 2.0, and adjust the space penalty item accordingly. The actual loss value is calculated by adding a small random perturbation of ±0.1% to ±2% to the theoretical value. The specific data of the iteration number and loss value are shown in Table 1.

[0024] Table 1. Data Statistics Table In this data analysis, a systematic evaluation was conducted on 30 iterations of the gradient descent correction mechanism during the optimization control of high-pressure jet grouting pile construction. The data table includes the strength deviation, low saturation node ratio, spatial penalty term, and loss value for each iteration. The data settings are intended to reflect the joint response characteristics of the loss function to the strength deviation and the uneven distribution of grout saturation. The theoretical construction of the loss value follows the basic logic of "square of intensity deviation multiplied by spatial penalty term". The spatial penalty term consists of 1 plus 2 times the proportion of low-saturation nodes, where the coefficient 2 is a preset spatial penalty coefficient. When the proportion of low-saturation nodes is 0, the spatial penalty term is 1, and the loss value is determined only by the square of the intensity deviation. As the proportion of low-saturation nodes increases, the spatial penalty term increases linearly, which has an amplification effect on the loss value under the same intensity deviation. The actual recorded loss value is superimposed with a small random disturbance on the basis of theoretical calculation, so that the data points do not fall completely on the theoretical surface, which truly reflects the uncertainty caused by numerical calculation and measurement noise in the parameter optimization process. By analyzing the data, we can observe the nonlinear coupling effect of intensity deviation and the proportion of low-saturation nodes on the loss value. For example, in iteration number 11, the intensity deviation is -1.108, and the absolute value of the deviation is small, but the proportion of low-saturation nodes is 0.331, the spatial penalty term is 1.662, and the final loss value is 2.041. This shows that even with a small deviation, a certain proportion of insufficient saturation will still increase the loss value. In iteration number 4, the intensity deviation reaches 3.108, the proportion of low-saturation nodes is 0.639, the spatial penalty term is as high as 2.278, and the loss value reaches 21.989. This indicates that when high intensity deviation and a high proportion of insufficient saturation work together, the loss value is amplified sharply. Conversely, in iteration number 21, the intensity deviation is only -0.345. Although the proportion of low-saturation nodes is 0.489, the spatial penalty term is 1.978, and the loss value is only 0.236. This shows that when the intensity deviation is close to zero, even if the saturation distribution is not very uniform, the loss value is still very low, and the optimization objective is mainly based on the intensity deviation. Regarding the amplification effect of the spatial penalty term, the intensity deviation of iteration number 14 is 1.874, and the intensity deviation of iteration number 25 is -2.145. The absolute values ​​of the deviations are similar, but the proportion of low-saturation nodes in the former is 0.829, while that in the latter is 0.359, resulting in spatial penalty terms of 2.658 and 1.718, respectively, and final loss values ​​of 9.341 and 7.904, respectively. This indicates that under similar intensity deviations, the more uneven the slurry filling, the higher the loss value, and the gradient descent correction mechanism will exert a stronger correction driving force. The intensity deviation of iteration number 3 is -0.457, the proportion of low-saturation nodes is only 0.092, the spatial penalty term is 1.184, and the loss value is only 0.247, which is one of the lowest losses among all iterations. This reflects that when both the intensity deviation and the proportion of insufficient saturation are low, the current parameter configuration is close to the optimal level. The presence of random perturbations causes slight deviations between the loss values ​​of some data points and the theoretical calculation values. For example, the intensity deviation of iteration number 20 is 2.678, the proportion of low-saturation nodes is 0.863, the theoretical loss value is about 19.48, and the actual recorded value is 19.556; the intensity deviation of iteration number 29 is -3.356, the proportion of low-saturation nodes is 0.523, the theoretical loss value is about 22.95, and the actual recorded value is 23.047. These small differences ensure that the fitted curve is "close to but not completely fitted" to the data points. When plotting the three-dimensional surface diagram of intensity deviation - proportion of low-saturation nodes - loss value, the data points will be distributed near the theoretical surface, forming a natural fluctuating distribution, which more realistically simulates the non-ideal characteristics in actual construction data collection and model calculation. In summary, the magnitude of the loss value is jointly determined by the strength deviation and the proportion of low-saturation nodes, and there is a non-linear multiplicative effect between the two. Data points with a combination of high deviation and high proportion will be assigned extremely high loss values, thus receiving priority correction in gradient descent. When the strength deviation is close to zero, even if there are certain defects in the saturation distribution, its optimization priority is relatively low. This design enables the construction parameter optimization process to quickly eliminate depth layers with substandard strength while also taking into account the uniformity of grout filling and avoiding the formation of local weak interlayers.

[0025] Calculate the partial derivative of the loss value with respect to the injection pressure in the initial boundary value to obtain the injection pressure gradient; calculate the partial derivative of the loss value with respect to the slurry flow rate in the initial boundary value to obtain the slurry flow rate gradient. A partial derivative is the derivative of a multivariable function with respect to one variable and the other variables treated as constants. The injection pressure gradient represents the instantaneous rate of change of the loss value with respect to changes in injection pressure when the slurry flow rate remains constant. The sign of the injection pressure gradient indicates whether the loss value increases or decreases with increasing injection pressure, and the absolute value of the injection pressure gradient indicates the sensitivity of the loss value to changes in injection pressure. Similarly, the slurry flow rate gradient represents the instantaneous rate of change of the loss value with changes in slurry flow rate when the injection pressure remains constant. The sign of the slurry flow rate gradient indicates whether the loss value increases or decreases with increasing slurry flow rate, and the absolute value of the slurry flow rate gradient indicates the sensitivity of the loss value to changes in slurry flow rate. The specific calculation process of the above partial derivatives is as follows: Based on the known predicted value of pile compressive strength, spatial distribution matrix of grout saturation and soil type code, the chain rule is used to calculate the derivative of the loss value with respect to the predicted value of pile compressive strength, the derivative of the predicted value of pile compressive strength with respect to the grout saturation of each grid node, and the derivative of the grout saturation of each grid node with respect to the initial values ​​of injection pressure and grout flow rate in the boundary. Through layer-by-layer back propagation, the values ​​of injection pressure gradient and grout flow rate gradient are finally obtained. Based on the soil type code of the target depth layer, the corresponding injection pressure response coefficient and grout flow rate response coefficient are determined by looking up a table from the pre-established correspondence between soil type codes and response coefficients. Soil category coding for target depth layer Determine two independent response coefficients: , , This is the injection pressure response coefficient, a positive real number, used to adjust the injection pressure update step size. This is the slurry flow response coefficient, a positive real number, used to adjust the slurry flow update step size; The correspondence between the above soil category codes and response coefficients is established in advance in the following way: select multiple sets of historical construction data belonging to the same soil category code, change the injection pressure and grout flow rate respectively, observe the change in the predicted value of pile compressive strength, calculate the ratio of the strength change to the parameter change as the sensitivity, take the reciprocal of the sensitivity or calibrate based on the sensitivity to obtain the response coefficient corresponding to the soil category code; In highly permeable soils, the intensity change caused by a unit change in grout flow rate is relatively small, requiring a larger response coefficient to accelerate convergence; in less permeable soils, the intensity change caused by a unit change in grout flow rate is relatively large, requiring a smaller response coefficient to prevent excessive parameter correction and oscillation. The response coefficient is a positive real number parameter pre-set according to the soil type code of the target depth layer, used to adjust the gradient descent correction step size. Since different soil types have different sensitivities to jet pressure and grout flow rate, in highly permeable sandy soil layers, the change in grout flow rate has a relatively small impact on grout saturation, requiring a larger response coefficient to accelerate convergence; in less permeable cohesive soil layers, the change in grout flow rate has a relatively large impact on grout saturation, requiring a smaller response coefficient to avoid over-correction of parameters. The pre-established correspondence between soil type codes and response coefficients is determined by performing sensitivity analysis or inversion analysis of historical construction data for different soil types. This correspondence is stored in the form of a lookup table, containing the jet pressure response coefficient and grout flow rate response coefficient corresponding to each soil type code. The injection pressure gradient is multiplied by the injection pressure response coefficient, and the product is defined as the injection pressure update step size. The injection pressure update step size is a positive or negative real number, the sign of which is determined by the direction of the injection pressure gradient, and its absolute value is determined by both the absolute value of the injection pressure gradient and the injection pressure response coefficient. The injection pressure update step size represents the correction direction and amount of the injection pressure in this iteration. The slurry flow rate gradient is multiplied by the slurry flow rate response coefficient, and the product is defined as the slurry flow rate update step size. The slurry flow rate update step size represents the correction direction and amount of the slurry flow rate in this iteration. Define the injection pressure gradient and slurry flow gradient : in, The injection pressure in the current initial boundary value, The slurry flow rate in the current initial boundary value; The gradient is calculated using the numerical perturbation method, taking injection pressure as an example: , For a preset small disturbance (such as The calculation method for the slurry flow gradient is the same as that for the current value of slurry flow (multiplied by 1). The numerical perturbation method avoids the analytical non-differentiability problem caused by the step characteristics of the saturation insufficient identification matrix, and is compatible with the black box characteristics of the intensity mapping program. Subtract the injection pressure update step from the current injection pressure, and subtract the slurry flow update step from the current slurry flow rate to complete the iterative update of the boundary initial value. During the iterative update process, the drill rod lifting speed and drill rod rotation speed in the boundary initial value always remain constant at the original value of the preset initial construction parameter vector. When the injection pressure gradient is positive, the injection pressure update step size is positive. After subtracting, the injection pressure decreases, thus reducing the loss value. When the injection pressure gradient is negative, the injection pressure update step size is negative. After subtracting, the injection pressure increases, which also reduces the loss value. This operation realizes the update of injection pressure along the negative gradient direction, so that the loss value gradually decreases along the direction of the fastest decreasing speed. Subtract the slurry flow rate update step size from the current slurry flow rate to obtain the updated slurry flow rate value. The update principle is the same as the update principle of the injection pressure. Replace the original injection pressure and slurry flow rate in the current layer boundary initial value with the updated injection pressure value and slurry flow rate value to complete one iteration update of the boundary initial value. The process of updating the step size and parameter iteration includes: , , , ,in, , These are the update steps for injection pressure and slurry flow rate, respectively. , The initial boundary value for the current iteration step. , Substitute the updated initial boundary values ​​into the next iteration step; During the iterative update process, the drill pipe lifting speed and drill pipe rotation speed remain unchanged from their original values ​​in the preset initial construction parameter vector; in the above formula, the strength deviation... The spatial distribution matrix of grout saturation is calculated from the predicted compressive strength of the pile output by S4 and the preset target strength value. The output, obtained from the steady-state solution obtained using the slurry diffusion model in S4, includes the soil category code. Obtained from S1, preset saturation threshold and space penalty coefficient The response coefficient is a constant pre-defined based on soil type and engineering experience. and These are predefined lookup table parameters; It should be noted that during the iterative update of the gradient descent correction mechanism, the drill rod lifting speed and drill rod rotation speed in the initial value of the current layer boundary always remain constant in the original value of the preset initial construction parameter vector and do not participate in gradient calculation and iterative update. The reason for this distinction is that in the actual high-pressure jet grouting construction process, the drill rod lifting speed and drill rod rotation speed mainly affect the construction efficiency and the uniformity of the pile diameter. They are conditional parameters that are predetermined in the construction preparation stage based on equipment capacity and engineering experience, and are not suitable for high-frequency dynamic adjustment layer by layer during the construction process. The injection pressure and grout flow rate directly affect the diffusion range and injection volume of the grout in the formation, and are active control parameters that play a decisive role in the strength of the pile. The two are controlled in real time and precisely by adjusting the output pressure of the injection pump and the output flow rate of the grout pump. By limiting the optimization variables to injection pressure and grout flow rate, the calculation dimension of the gradient descent correction mechanism is reduced from four dimensions to two dimensions. The convergence direction of the optimization process is precisely matched with the actual adjustable degrees of freedom of the construction equipment. The obtained correction results can be directly converted into equipment control commands, avoiding the disconnect between the optimization results and the actual on-site operation.

[0026] S6: Repeat steps S2 to S5 to update the initial boundary values ​​of the slurry diffusion model until the deviation converges, and obtain the optimal construction parameters of the current depth layer. Only the slurry flow rate and injection pressure are extracted as the initial boundary values ​​for the next adjacent depth layer iteration calculation. The initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, thus obtaining the optimal construction parameters for the current depth layer. The specific steps are as follows: Repeatedly updating the initial boundary values ​​of the grout diffusion model until the deviation converges means using the initial boundary values ​​updated by the gradient descent correction mechanism as the input for a new round of calculations, and iteratively executing the construction and solution of the grout diffusion model, the generation of the grout saturation spatial distribution matrix, the calculation of the strength mapping relationship model, and the deviation comparison of the predicted pile compressive strength values, forming an iterative closed loop until the deviation between the predicted pile compressive strength value and the preset target strength value meets the convergence condition. Iteration refers to using the updated initial boundary value of the previous calculation as the input initial boundary value for the next calculation, and repeating the same calculation process. Each complete calculation process is called an iteration. The purpose of iteration is to gradually correct the injection pressure and grout flow rate in the initial boundary value so that the predicted value of the pile compressive strength continuously approaches the preset target strength value until the deviation is reduced to an acceptable range. Reconstructing the slurry diffusion model based on the iteratively updated initial boundary values ​​refers to replacing the corresponding parameters in the original initial boundary values ​​with the updated injection pressure and slurry flow rate values ​​after the gradient descent correction mechanism. These parameters are then recombined with the soil type code, permeability coefficient, and drill rod lifting speed and drill rod rotation speed in the preset initial construction parameter vector at the target depth layer. The slurry phase continuity equation, slurry phase seepage equation, and soil skeleton deformation equation are then re-established in the manner described in step S3, forming a new set of fluid-structure interaction control equations. The purpose of the reconstruction is to make the slurry diffusion model reflect the construction parameter conditions after the initial boundary values ​​are updated. Calculating the dynamic pressure field and slurry saturation spatial distribution matrix by reconstructing the slurry diffusion model refers to solving the fluid-structure interaction control equations formed after reconstruction to obtain the slurry pressure value and soil displacement value of each grid node under the updated boundary initial conditions. Based on the slurry pressure value, the slurry saturation value of each grid node is calculated. After the slurry pressure field and soil displacement field reach a steady state, the slurry saturation value of each grid node is arranged according to the spatial position of the grid node to form a new slurry saturation spatial distribution matrix. The new predicted value of pile compressive strength is obtained by processing the spatial distribution matrix of grout saturation through a strength mapping relationship model. This involves combining the grout saturation value of each grid node in the new spatial distribution matrix with the soil category code of the target depth layer, inputting it into a pre-calibrated strength mapping relationship model, obtaining the median value of pile compressive strength for each grid node, weighting the median value of pile compressive strength for each grid node according to the preset weight coefficient of the soil category code corresponding to each grid node, obtaining the weighted value of pile compressive strength for each grid node, and calculating the average value of the weighted values ​​of pile compressive strength for all grid nodes within the target depth layer to obtain the new predicted value of pile compressive strength. The deviation between the new predicted value of pile compressive strength and the preset target strength value is calculated by subtracting the preset target strength value from the predicted value of pile compressive strength obtained in this iteration. The absolute value of the deviation reflects whether the construction parameters corresponding to the initial boundary value after this iteration have met the design requirements. The iteration stops when the absolute value of the deviation is less than or equal to the preset convergence threshold. The preset convergence threshold is a pre-set upper limit of the absolute value of the deviation, used to determine whether the iterative process has reached the convergence condition. The preset convergence threshold is a positive real number, and its specific value is determined according to the engineering accuracy requirements, set as a certain percentage of the preset target strength value. The smaller the value of the preset convergence threshold, the higher the accuracy of the optimization result, but the more iterations are required; the larger the value, the faster the iteration convergence speed, but the greater the deviation between the optimization result and the target strength. When the absolute value of the deviation is less than or equal to the preset convergence threshold, it indicates that the predicted value of the pile compressive strength under the current boundary initial value condition is close enough to the preset target strength value. The correction range of the boundary initial value will be very limited if the iteration continues. At this time, the deviation is judged to converge and the iteration is stopped. Deviation convergence means that the absolute value of the deviation is reduced to below the preset convergence threshold for the first time during the iteration process. Reaching deviation convergence means that the optimization of the construction parameters of the current depth layer has been completed. The injection pressure and grout flow rate at the point where iteration stops are determined as the optimal injection pressure and optimal grout flow rate for the target depth layer. The optimal injection pressure refers to the injection pressure value that, under the current geological conditions and construction parameter configuration of the target depth layer, enables the predicted pile compressive strength to meet the preset target strength value and is confirmed through iterative convergence. The optimal grout flow rate refers to the grout flow rate value that, under the current geological conditions and construction parameter configuration of the target depth layer, enables the predicted pile compressive strength to meet the preset target strength value and is confirmed through iterative convergence. The drill pipe lifting speed in the preset initial construction parameter vector is determined as the optimal drill pipe lifting speed for the target depth layer. Throughout the entire iterative optimization process of the target depth layer, the drill pipe lifting speed remains unchanged in the original value in the preset initial construction parameter vector and does not participate in the iterative update of the gradient descent correction mechanism. Therefore, the optimal drill pipe lifting speed is directly equal to the drill pipe lifting speed in the preset initial construction parameter vector. The drill pipe rotation speed in the preset initial construction parameter vector is determined as the optimal drill pipe rotation speed for the target depth layer. Similarly to the drill pipe lifting speed, the drill pipe rotation speed also remains unchanged in the original value in the preset initial construction parameter vector throughout the entire iterative optimization process of the target depth layer, and does not participate in the iterative update of the gradient descent correction mechanism. Therefore, the optimal drill pipe rotation speed is directly equal to the drill pipe rotation speed in the preset initial construction parameter vector. The optimal construction parameters for the target depth layer consist of the optimal injection pressure, optimal slurry flow rate, optimal drill rod lifting speed, and optimal drill rod rotation speed. The optimal construction parameters are a parameter set containing four components, which correspond to the four key operational variables that need to be controlled during the high-pressure jet grouting pile construction process. Among them, the optimal injection pressure and optimal slurry flow rate are converged values ​​obtained through iterative optimization, while the optimal drill rod lifting speed and optimal drill rod rotation speed are the original values ​​in the preset initial construction parameter vector. The optimal construction parameters represent the optimal combination of parameters that can ensure construction efficiency while meeting the preset target strength value requirements under the current geological conditions of the target depth layer. Only the optimal slurry flow rate and optimal injection pressure from the optimal construction parameters are extracted as the initial boundary values ​​for the iterative calculation of the next adjacent depth layer. The next adjacent depth layer refers to the depth layer immediately below the current target depth layer in the depth direction. After the current target depth layer is optimized, its optimal slurry flow rate value and optimal injection pressure value are extracted and used as the initial values ​​of slurry flow rate and injection pressure in the initial boundary values ​​of the next adjacent depth layer, respectively. The physical significance of this extraction and transfer operation lies in the fact that high-pressure jet grouting pile construction is a continuous process that proceeds from top to bottom or from bottom to top along the depth direction. When the construction of the previous depth layer ends, the actual grout flow rate and injection pressure at the nozzle directly determine the initial injection state of the grout at the interface of the adjacent depth layer. Transferring these two parameters to the next depth layer as the initial boundary value can ensure the interlayer continuity of the entire pile construction parameter sequence from a physical mechanism perspective.

[0027] S7: Switch the target depth layer layer by layer along the depth direction of the pile body, repeat steps S2 to S6 until all depth layers are traversed, and generate the optimal construction parameter sequence of the whole pile according to the depth order of each depth layer. The target depth layer is switched layer by layer along the depth direction of the pile until all depth layers of the entire pile have been traversed. The specific steps are as follows: Switching the target depth layer layer by layer along the depth direction of the pile body means that each depth layer is determined as the current target depth layer to be optimized in the order of shallow to deep. After the optimization process is completed in the current target depth layer and the optimal construction parameters of the layer are obtained, the optimization object is switched to the next adjacent depth layer until the bottom depth layer of the pile body is optimized. The layer-by-layer switching method ensures that each depth layer is optimized independently under the constraints of the stratum conditions and construction parameters. At the same time, the construction status between adjacent depth layers is continuously connected through the transfer mechanism of the initial value of the interlayer boundary. The first layer is taken as the current depth layer. The first layer refers to the shallowest layer among several continuous depth layers divided along the depth direction of the pile body, which corresponds to the starting depth range of the pile top. The current depth layer refers to the target depth layer that is currently being optimized. During the layer-by-layer switching process, the specific depth layer pointed to by the current depth layer moves downward as the optimization process progresses. The preset initial grout flow rate and preset initial injection pressure are used as the initial boundary values ​​of the current depth layer. The preset initial grout flow rate is the initial grout flow rate value of the first layer determined by engineering experience, equipment rated discharge capacity and the type of the first layer soil during the construction preparation stage. The preset initial injection pressure is the initial injection pressure value of the first layer determined by engineering experience, equipment rated pressure and the type of the first layer soil during the construction preparation stage. These two preset values ​​are used as the grout flow rate and injection pressure in the initial boundary values ​​of the first layer, providing initial calculation conditions for the construction of the grout diffusion model of the first layer and subsequent iterative optimization. The gradient descent correction mechanism is executed and the initial boundary values ​​of the grout diffusion model are repeatedly updated until the deviation converges, thereby obtaining the optimal construction parameters for the first layer. All operations described in steps S2 to S6 are performed sequentially on the first layer, including constructing a grout diffusion model with preset initial grout flow rate and initial injection pressure as initial boundary values, calculating the dynamic pressure field and the spatial distribution matrix of grout saturation, obtaining the predicted value of pile compressive strength through the strength mapping relationship model, calculating the deviation and determining whether it exceeds the limit, activating the gradient descent correction mechanism to iteratively update the initial boundary values ​​when the deviation exceeds the limit, and repeating this iterative process until the deviation converges to within the preset convergence threshold, thereby finally determining the optimal injection pressure, optimal grout flow rate, optimal drill rod lifting speed, and optimal drill rod rotation speed for the first layer. Extract the optimal slurry flow rate and optimal injection pressure of the current depth layer and use them as the initial values ​​of the iteration boundary of the next depth layer. The current depth layer is the first layer, and the next depth layer is the second depth layer immediately below the first layer according to the depth arrangement order. Take the optimal slurry flow rate value of the first layer and use it as the initial value of the slurry flow rate in the initial value of the boundary of the second depth layer. Take the optimal injection pressure value of the first layer and use it as the initial value of the injection pressure in the initial value of the boundary of the second depth layer. The initial value of the iteration boundary refers to the initial boundary value used by the next depth layer when performing the iterative optimization process. This value comes from the optimal construction parameters after the optimization convergence of the previous depth layer and is different from the preset initial value used by the first layer. The layer is changed according to the depth arrangement order, and the next depth layer is determined as the new current depth layer. The depth arrangement order refers to the natural arrangement order of the depth layers from shallow to deep, with the first layer being the first layer, and the second layer, the third layer and so on down to the pile bottom depth layer. The layer change refers to switching the optimization object from the completed depth layer to the next depth layer that has not yet been optimized, and updating the direction of the current depth layer. Repeatedly execute the gradient descent correction mechanism and repeatedly update the initial boundary values ​​of the slurry diffusion model until the deviation converges to obtain the optimal construction parameters for the depth layer. For the updated current depth layer, using its iterative boundary initial values ​​as the starting condition, repeat all optimization processes from steps S2 to S6 to obtain the optimal injection pressure, optimal slurry flow rate, optimal drill pipe lifting speed, and optimal drill pipe rotation speed for the depth layer. After the optimization of the depth layer is completed, extract its optimal slurry flow rate and optimal injection pressure again as the initial iterative boundary values ​​for the next depth layer. The optimization process is carried out step by step, iterating until the parameters of the last layer at the bottom of the pile are optimized. Step by step, iterating means that after the optimization of each depth layer is completed, the optimal grout flow rate and optimal injection pressure in the optimization results are passed to the next adjacent depth layer. The next depth layer continues to execute the optimization process based on the optimization results of the previous depth layer, forming a step-by-step iterative chain. The last layer at the bottom of the pile refers to the deepest layer in a series of continuous depth layers divided along the depth direction of the pile body, which corresponds to the termination depth range of the pile bottom. When the gradient descent correction mechanism and deviation convergence judgment of the last layer at the bottom of the pile are completed and the optimal construction parameters of the layer are obtained, the optimization of construction parameters for all depth layers of the entire pile is completed. The optimal construction parameters for all depth layers are arranged in depth order to generate the optimal construction parameter sequence for the entire pile. The depth order refers to the order in which the depth layers are arranged from shallowest to deepest, that is, from the first layer to the last layer at the bottom of the pile. The optimal construction parameters for the first layer are placed in the first position of the sequence, the optimal construction parameters for the second layer are placed in the second position, and so on, until the optimal construction parameters for the last layer at the bottom of the pile are placed in the last position of the sequence. The optimal construction parameter sequence for the entire pile is a four-dimensional parameter sequence arranged in depth order, containing the optimal injection pressure, optimal grout flow rate, optimal drill rod lifting speed, and optimal drill rod rotation speed for each depth layer. In this sequence, the b-th group of parameters corresponds to the b-th depth layer. The construction equipment calls and executes the parameters segment by segment during the construction process according to the depth interval corresponding to each group of parameters in this sequence, thereby achieving precise segmented control of the construction parameters within the entire pile length. Through the above-mentioned step-by-step iterative optimization method, the actual grout injection state and jetting energy level at the end of the construction of the previous depth layer are transferred to the initial calculation conditions of the next depth layer through the two parameters of optimal grout flow rate and optimal jetting pressure. This establishes a continuity constraint in the physical mechanism between adjacent depth layers, so that there is no physically unrealizable step jump between the construction parameters of adjacent depth layers in the final generated optimal construction parameter sequence of the whole pile. This avoids the technical problem of sudden changes in pile strength along the depth direction or the formation of weak interlayers caused by discontinuity between parameter sequence layers.

[0028] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0029] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0030] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0031] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A high-pressure jet grouting pile construction optimization control method, characterized in that, The specific steps include: S1: Divide the target pile location into several continuous depth layers, obtain the soil type code and permeability coefficient of each depth layer, and determine the target depth layer to be optimized. S2: Use the slurry flow rate and injection pressure of the adjacent depth layer above the target depth layer as the initial value of the current layer boundary. If the current depth layer is the first layer, then use the preset initial slurry flow rate and initial injection pressure as the initial value of the boundary. S3: Combining the soil type, permeability coefficient, initial boundary value, and initial construction parameter vector of the target depth layer, construct a grout diffusion model based on fluid-structure interaction theory; S4: Use the grout diffusion model to solve the dynamic pressure field of the grout inside the soil at the target depth layer, generate the spatial distribution matrix of grout saturation based on the dynamic pressure field, input the spatial distribution matrix into the pre-calibrated strength mapping relationship model, and perform weighted calculations in combination with the soil category code of the target depth layer to obtain the predicted value of the pile compressive strength. S5: Compare the predicted value of the pile compressive strength with the preset target strength value. If the deviation exceeds the limit, activate the gradient descent correction mechanism to dynamically correct the initial value of the boundary of the current depth layer. S6: Repeat steps S2 to S5 to update the initial boundary values ​​of the slurry diffusion model until the deviation converges, and obtain the optimal construction parameters of the current depth layer. Only the slurry flow rate and injection pressure are extracted as the initial boundary values ​​for the next adjacent depth layer iteration calculation. S7: Switch the target depth layer layer by layer along the depth direction of the pile body, repeat steps S2 to S6 until all depth layers are traversed, and generate the optimal construction parameter sequence for the whole pile according to the depth order of each depth layer. 2.The high-pressure jet grouting pile construction optimization control method of claim 1, wherein: The specific steps to obtain the soil category code and permeability coefficient for each depth layer are as follows: The target pile location is divided into several continuous depth layers at equal intervals according to a preset fixed layer thickness along the depth direction of the pile location. In-situ static cone penetration test and standard penetration test were conducted on the geological exploration borehole at the target pile location to obtain cone tip resistance, side friction resistance and standard penetration blow count; For each depth layer, the soil type code is determined based on the cone tip resistance, side friction resistance, and standard penetration test blow count, and the permeability coefficient of the depth layer is calibrated by field borehole water injection test or variable head permeability test.

3. The high-pressure jet grouting pile construction optimization control method according to claim 2, characterized in that: The method for determining soil category codes is as follows: A ratio-type parameter is constructed to characterize the particle characteristics of soil. The ratio-type parameter is determined by the cone tip resistance and the side friction resistance. A set of boundary values ​​for soil category classification is predefined. The set of boundary values ​​includes at least the upper and lower threshold values ​​for ratio parameters, as well as the upper and lower threshold values ​​for standard penetration test blow counts. Based on the comparison results of the ratio class parameters and the standard penetration test blow count relative to the limit value, the preset soil category code corresponding to the soil at this depth layer is determined by matching; If the conditions for determining any of the preset soil category codes are not met, the soil at that depth layer will be classified as the preset catch-all soil category code.

4. The high-pressure jet grouting pile construction optimization control method of claim 1, wherein: The slurry diffusion model based on fluid-structure interaction theory is constructed and used to output the spatial distribution matrix of slurry saturation at the target depth layer in the following manner: The computational domain of the target depth layer is discretized using the finite volume method, resulting in a discrete grid composed of multiple grid nodes, with each grid node corresponding to a spatial location within the computational domain. The injection pressure in the initial value of the target depth layer boundary is used as the slurry pressure boundary condition of the grid node where the nozzle is located. The drill rod lifting speed and rotation speed in the initial construction parameter vector are used to set the nozzle depth movement rate and circumferential rotation rate, respectively. The absolute permeability of the soil is obtained by converting the permeability coefficient of the target depth layer, and the relative permeability of the grout corresponding to the target depth layer is determined from the pre-established correspondence between the soil category code and the relative permeability of the grout. Establish the slurry phase continuity equation, slurry phase seepage equation, and soil skeleton deformation equation to form a slurry diffusion model based on fluid-structure interaction theory; Solve the fluid-structure interaction control equations simultaneously to obtain the grout pressure and soil displacement at each grid node, and calculate the grout saturation at each grid node based on the grout pressure. When the changes in grout pressure and soil displacement of all grid nodes are less than the corresponding preset convergence thresholds, the calculation is determined to have reached a steady state. The grout saturation of each grid node under steady state is arranged according to the spatial position of the grid node to form a spatial distribution matrix of grout saturation.

5. The optimized control method for high-pressure jet grouting pile construction according to claim 4, characterized in that: The spatial distribution matrix of grout saturation is input into a pre-calibrated strength mapping model, and weighted calculations are performed in conjunction with the soil category code of the target depth layer to obtain the predicted value of pile compressive strength. The specific steps are as follows: Collect multiple sets of historical high-pressure jet grouting pile engineering data, and obtain the measured values ​​of pile core compressive strength, grout saturation value and soil type code for each grid node at different depth layers; Using the slurry saturation value and soil type code at each grid node as input, and the measured core compressive strength of the pile as output, a strength mapping relationship model under different soil type codes is constructed. Extract the slurry saturation of each grid node in the spatial distribution matrix of slurry saturation, and combine it with the soil category code of the target depth layer to input the strength mapping relationship model. Define the pile compressive strength calculation result output by the strength mapping relationship model for each grid node as the intermediate value of pile compressive strength. Based on the preset weight coefficients of the soil category codes corresponding to each grid node, the median value of the pile compressive strength of each grid node is weighted to obtain the weighted value of the pile compressive strength of each grid node. The weighted average value of the pile compressive strength of all grid nodes within the target depth layer is calculated to obtain the predicted value of the pile compressive strength of the target depth layer.

6. The method for optimizing and controlling the construction of high-pressure jet grouting piles according to claim 5, characterized in that: The intensity mapping relationship model uses different function structures for coding different soil types; the function structures are selected from a pre-established set of function structures, which includes at least exponential functions, linear functions, and piecewise functions. The fitting parameters in each function were determined by performing regression analysis on historical data belonging to the same soil category code.

7. The method for optimizing and controlling the construction of high-pressure jet grouting piles according to claim 1, characterized in that: When the deviation exceeds the limit, the gradient descent correction mechanism is activated. The specific steps are as follows: The squared deviation between the predicted pile compressive strength value at the target depth layer and the preset target strength value is taken as the loss value; Calculate the partial derivative of the loss value with respect to the injection pressure in the initial boundary value to obtain the injection pressure gradient; calculate the partial derivative of the loss value with respect to the slurry flow rate in the initial boundary value to obtain the slurry flow rate gradient. Based on the soil type code of the target depth layer, the corresponding injection pressure response coefficient and grout flow rate response coefficient are determined by looking up a table from the pre-established correspondence between soil type codes and response coefficients. The injection pressure gradient is multiplied by the injection pressure response coefficient, and the product is defined as the injection pressure update step size. The slurry flow rate gradient is multiplied by the slurry flow rate response coefficient, and the product is defined as the slurry flow rate update step size. Subtract the injection pressure update step from the current injection pressure, and subtract the slurry flow update step from the current slurry flow rate to complete the iterative update of the boundary initial values. During the iterative update process, the drill rod lifting speed and drill rod rotation speed in the boundary initial values ​​always remain constant at the original values ​​of the preset initial construction parameter vector.

8. The optimized control method for high-pressure jet grouting pile construction according to claim 7, characterized in that: The initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, thus obtaining the optimal construction parameters for the current depth layer. The specific steps are as follows: Based on the iteratively updated initial boundary values, the grout diffusion model is reconstructed. The dynamic pressure field and the spatial distribution matrix of grout saturation are calculated using the reconstructed grout diffusion model. The spatial distribution matrix of grout saturation is then processed using a strength mapping relationship model to obtain new predicted values ​​of pile compressive strength. Calculate the deviation between the new predicted value of pile compressive strength and the preset target strength value; stop the iteration when the absolute value of the deviation is less than or equal to the preset convergence threshold; determine the injection pressure and grout flow rate at the time of stopping the iteration as the optimal injection pressure and optimal grout flow rate for the target depth layer; The drill rod lifting speed in the preset initial construction parameter vector is determined as the optimal drill rod lifting speed for the target depth layer; the drill rod rotation speed in the preset initial construction parameter vector is determined as the optimal drill rod rotation speed for the target depth layer; The optimal construction parameters for the target depth layer consist of the optimal injection pressure, optimal slurry flow rate, optimal drill pipe lifting speed, and optimal drill pipe rotation speed.

9. The method for optimizing and controlling the construction of high-pressure jet grouting piles according to claim 8, characterized in that: The target depth layer is switched layer by layer along the depth direction of the pile until all depth layers of the entire pile have been traversed. The specific steps are as follows: The first layer is taken as the current depth layer. The preset initial slurry flow rate and preset initial injection pressure are used as the initial boundary values ​​of the current depth layer. The gradient descent correction mechanism is executed and the initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, so as to obtain the optimal construction parameters of the first layer. Extract the optimal slurry flow rate and optimal injection pressure of the current depth layer, and use them as the initial values ​​of the iteration boundary of the next depth layer; The layers are iterated according to the depth arrangement order, the next depth layer is determined as the new current depth layer, the gradient descent correction mechanism is repeatedly executed and the initial boundary values ​​of the slurry diffusion model are repeatedly updated until the deviation converges, and the optimal construction parameters of the depth layer are obtained. The process is iterated step by step until the parameters of the last layer at the bottom of the pile are optimized. The optimal construction parameters for all depth layers are arranged in order of depth to generate a sequence of optimal construction parameters for the entire pile.

Images