Edge computing-based building engineering grouting operation collaborative control method and system

Abstract

This invention discloses a collaborative control method and system for grouting operations in building engineering based on edge computing, belonging to the field of grouting control technology. The invention collects construction data, grouting parameters, and engineering parameters; performs Kalman filtering state estimation on each grouting hole to obtain an estimated diffusion distance of the grout diffusion front at the current moment; performs spatial geometric partitioning and interference quantization, calculating the control area and geometric interference area using weighted Thiessen polygons; uses a multi-objective evolution algorithm to solve for the optimal combination of grouting pressure and grout dynamic viscosity for the next control cycle, with the optimization objectives of maximizing the sealing efficiency of the corrected fracture section and maximizing the strength of the time-cured structure; performs stress intensity factor safety verification on the optimal parameter combination, and sends out the parameters that pass the verification for execution. This invention achieves closed-loop collaborative optimization of multi-hole grouting, effectively avoiding the risks of grout cross-contamination and secondary formation fracturing, and significantly improving grouting sealing efficiency and the uniformity of the solidified body strength.

Description

Technical Field

[0001] This invention relates to the field of grouting control technology, and in particular to a collaborative control method and system for grouting operations in building engineering based on edge computing. Background Technology

[0002] Grouting is a core technical means for ground reinforcement, seepage prevention, and disaster management in underground space construction such as building engineering and mining. Due to the heterogeneity and complexity of underground geological media, it is necessary to precisely control key parameters such as grouting pressure, flow rate, and grout properties during the grouting process to ensure that the grout can effectively diffuse and fill the fracture space, while achieving the expected engineering curing strength. Chinese patent application CN121300219A discloses a grouting system and its control method. The technical solution uses a programmable logic controller as the core control unit. By systematically integrating key equipment such as grouting pumps and valve groups, an intelligent equipment cluster is built. With the help of a distributed sensor network, key operating parameters such as pressure, flow rate, and concentration are monitored and collected in real time. In terms of control logic, the solution introduces a PID controller to control the grouting pressure and optimizes the fuzzy control of the PID controller based on the sparrow search algorithm. By monitoring the flow status of the grouting pump in real time, the control coefficient is updated adaptively, thereby adjusting the motor speed and the opening and closing degree of the solenoid valve. The aforementioned existing technical solutions mainly focus on the logic control of a single device, without deeply considering the dynamic spatial topological relationship and geometric interference effect of the grout diffusion front of adjacent grouting holes in multi-hole operations. In multi-hole collaborative grouting operations, if the grout diffusion front of adjacent grouting holes is excessively compressed, there is a real risk that the grouting medium will undergo secondary splitting. Existing technologies lack quantification and avoidance mechanisms for such spatial geometric interference. Since the true diffusion boundary of the grout in the building foundation medium is not visible, existing technologies rely only on surface monitoring data at the borehole opening for passive feedback adjustment, lacking real-time dynamic estimation of internal states such as the radial diffusion distance of the underground grout. This results in the control system being unable to accurately map and verify the actual crack section sealing effectiveness. In addition, existing technologies only adjust single variables such as pressure, failing to combine grouting pressure and grout dynamic viscosity for multi-variable and multi-objective collaborative optimization, making it difficult to maximize the time-cured structural strength of the grout while ensuring the grout diffusion range. Summary of the Invention

[0003] The technical problem solved by this invention is that existing technologies are difficult to use for multi-objective coordinated control of pressure and viscosity based on spatial geometric topological interference and internal diffusion state estimation in multi-hole coordinated grouting operations, and lack physical safety prevention mechanisms based on fracture mechanics, which easily leads to problems such as grout cross-contamination between holes, secondary fracturing of the formation, and substandard overall sealing and solidification performance during construction.

[0004] To address the aforementioned technical problems, this invention provides the following technical solution: a collaborative control method for grouting operations in building engineering based on edge computing, comprising the following steps: Step S1: Collect construction data, grouting parameters, and engineering parameters; Step S2: Based on the construction data, grouting parameters, and engineering parameters, perform a state estimation algorithm on each grouting hole to obtain an estimated value of the diffusion distance of the grout diffusion front at the current moment; Step S3: Based on the construction data, perform spatial geometric subdivision and interference quantization to calculate the control area and geometric interference area of ​​each grouting hole; Step S4: Based on the estimated values ​​of the control area, geometric interference area, and diffusion distance, a multi-objective optimization algorithm is used to solve for the optimal combination of grouting parameters for the next control cycle. Step S5: Perform a safety verification on the optimal grouting parameter combination, and send out the parameters that pass the verification for execution.

[0005] Preferably, in step S1, the construction data includes the grouting pressure and grouting flow rate during the grouting process; The grouting parameters include the set value of the dynamic viscosity of the grout. The engineering parameters include: the permeability of the soil in the grouting area, the effective thickness of the grouting layer through which the grouting hole passes, the environmental pore water pressure, the effective porosity of the medium in the grouting area, the crack initiation toughness constant of the grouting medium, the radius of the grouting hole, and the design target diffusion radius.

[0006] Preferably, in step S2, the state estimation algorithm is a Kalman filter algorithm, and the execution process includes: The radial distance and radial diffusion velocity of the grout in each grouting hole are constructed as a state vector; At each sampling time, the prior predicted state vector and prediction error covariance matrix at the current time are calculated using the state transition equation based on the state vector at the previous time. Perform an observation validity determination, the logic of which includes: If the grouting pressure at the sampling moment is greater than the environmental pore water pressure, and the grouting flow rate is greater than or equal to the effective flow velocity threshold, then the grouting flow rate is converted into volumetric flow rate, substituted into the radial Darcy flow formula to obtain the observed diffusion radius as the observed value, and the observation matrix is ​​set; the Kalman gain matrix is ​​calculated; based on the Kalman gain matrix, the prior predicted state vector, the observed value, and the observation matrix, the posterior state estimation vector at the current moment is calculated, and the posterior estimation error covariance matrix is ​​updated; If the grouting pressure at the sampling time is less than or equal to the environmental pore water pressure, or the grouting flow rate is less than the effective flow velocity threshold, then the calculation of the observed value and Kalman gain is skipped, the radial diffusion velocity at the current time is forcibly set to 0, the radial distance is kept as the posterior state estimate of the previous time, and the posterior estimation error covariance matrix is ​​set to be equal to the posterior estimation error covariance matrix of the previous time plus the process noise covariance matrix; finally, the diffusion distance estimate of each grouting hole at each sampling time is output.

[0007] Preferably, in step S3, the process of spatial geometric subdivision and interferometric quantization specifically includes: At the beginning of each control cycle, obtain the estimated values ​​of grouting pressure and diffusion distance from the previous sampling time. Based on the grouting pressure of each grouting hole, a weighting factor is calculated. The process of calculating the weighting factor specifically includes: Calculate the maximum and minimum pressure values ​​among all grouting holes, calculate the normalized value for each grouting hole, and set the weighting factor to the normalized value plus 1; Using the planar coordinates of each grouting hole as seed points and weighting factors as weights, the multiplicative weighted Voronoi diagram algorithm is executed to generate the weighted Thiessen polygon region corresponding to each grouting hole; For any point on a two-dimensional plane, divide the actual geometric distance from the point to the grouting hole by the weighting factor of the grouting hole to obtain the weighted distance. Divide each point into the grouting hole that minimizes its weighted distance to obtain the weighted Thiessen polygon region corresponding to each grouting hole. Based on the ordered vertex coordinate sequence of the weighted Thiessen polygon region corresponding to each grouting hole, the area enclosed by the weighted Thiessen polygon region is calculated using the shoelace formula, which serves as the control area of ​​the grouting hole.

[0008] Preferably, the process of calculating the geometric interference area in step S3 specifically includes: In the interference calculation of the buffer warning zone, the weighted Thiessen polygon region of each grouting hole is shifted outward by a preset safe buffer distance to generate a buffer extended polygon. For all adjacent grouting hole pairs, calculate the area of ​​the intersection region of the two buffer extended polygons, where adjacent grouting hole pairs are grouting hole pairs with a common edge in the corresponding weighted Thiessen polygon regions; For each grouting hole, the geometric interference area of ​​the intersection region generated by it and all adjacent grouting holes is summed.

[0009] Preferably, in step S4, the parameter combination optimized by the multi-objective optimization algorithm is grouting pressure and grout dynamic viscosity; The optimization objectives include: maximizing the crack sealing efficiency and maximizing the strength of the age-cured structure; The process of calculating the maximum crack sealing efficiency specifically includes: Predict the grout diffusion distance after one control cycle under the action of candidate grouting pressure and grout dynamic viscosity; The crack sealing efficiency is calculated based on the ratio of the predicted slurry diffusion distance to the target diffusion radius; A penalty factor based on spatial geometric interference is introduced to correct the sealing effectiveness of the fracture section. The correction process specifically includes: The ratio of the geometric interference area to the control area of ​​each grouting hole is used as the spatial penalty factor, and the spatial penalty factor is truncated. The crack section sealing effectiveness is corrected based on the truncated spatial penalty factor to obtain the corrected crack section sealing effectiveness.

[0010] Preferably, the mathematical expression for the strength of the age-cured structure is: ; in, For age-cured structural strength, The first fitting coefficient, The dynamic viscosity of the candidate slurry. For reference viscosity, is the second fitting coefficient.

[0011] Preferably, in step S4, the evolution process of the multi-objective optimization algorithm specifically includes: The population size and number of iterations of the preset evolutionary algorithm are used to select samples from the parent population through a tournament selection method. Simulated binary crossover is performed on the samples with a set probability to generate offspring. Mutation operation is performed on all generated offspring individuals, wherein the mutation operation is determined by the current spatial topology. The process of performing the mutation operation specifically includes: Extract the spatial penalty factor value after the grouting hole is truncated and compare it with the preset penalty threshold; if the truncated spatial penalty factor value is less than the penalty threshold, it is determined that the current grouting hole is in a safe and cooperative state, and regular Gaussian mutation is performed. If the truncated space penalty factor value is greater than or equal to the penalty threshold, the current grouting hole is determined to be in a collision danger state. The directional dynamic compression of the variation space is performed, the upper limit of the allowable variation of the candidate grouting pressure is temporarily lowered, the lower limit of the allowable variation of the candidate grout dynamic viscosity is temporarily raised, and conventional Gaussian variation is performed within the compressed range. After the optimization algorithm completes its iterations, it outputs the Pareto front non-dominated solution set on the optimization objective. All non-dominated solutions on the Pareto front are sorted, and the solution ranked first is extracted as the candidate optimal parameter combination. The evaluation algorithm is called to calculate the stress intensity factor corresponding to the parameter combination. If the stress intensity factor is less than the crack initiation toughness constant of the grouting medium, the parameter combination is taken as the final safe and feasible solution. If the stress intensity factor is greater than or equal to the crack initiation toughness constant of the grouting medium, then the next non-dominated solution is extracted sequentially for verification until a parameter combination that satisfies the stress intensity factor being less than the crack initiation toughness constant of the grouting medium is found. The final grouting pressure and grout dynamic viscosity values ​​are sent to the corresponding grouting unit's programmable logic controller for execution.

[0012] Preferably, the evaluation algorithm takes the candidate grouting pressure, the continuous grouting time within the current control cycle, and the dynamic viscosity of the grout as inputs, calculates the stress intensity factor at the crack tip, and takes the stress intensity factor as the output. The evaluation algorithm process specifically includes: The cumulative injection volume of the grout under the action of candidate parameters is calculated, and the mathematical expression for the cumulative injection volume is: ; in, For the cumulative injection volume, For the comprehensive flow coefficient, For grouting pressure, For continuous grouting time, The dynamic viscosity of the slurry; The fracture half-length is calculated based on the cumulative injected volume, and the stress intensity factor is calculated based on the grouting pressure and the fracture half-length. The mathematical expression for the stress intensity factor is as follows: ; in, Stress intensity factor For grouting pressure, The length of the fissure is half its length.

[0013] A collaborative control system for grouting operations in building engineering based on edge computing includes a data acquisition module, an estimation module, a data segmentation module, an optimization module, and a verification module. The acquisition module is used to collect construction data, grouting parameters, and engineering parameters; The estimation module is used to perform a state estimation algorithm on each grouting hole based on the construction data, grouting parameters and engineering parameters, to obtain an estimated value of the diffusion distance of the grout diffusion front at the current moment; The segmentation module is used to perform spatial geometric segmentation and interference quantization based on the construction data, and to calculate the control area and geometric interference area of ​​each grouting hole; The optimization module is used to solve for the optimal grouting parameter combination for the next control cycle based on the estimated control area, geometric interference area, and diffusion distance using a multi-objective optimization algorithm. The verification module is used to perform safety verification on the optimal grouting parameter combination and send out the parameters that pass the verification for execution.

[0014] The beneficial effects of this invention are as follows: This invention utilizes the Kalman filter algorithm to estimate the radial distance of the grout diffusion front of each grouting hole in real time. Based on this, a weighted Thiessen polygon is dynamically generated based on the grouting pressure to quantify the effective control area of ​​each grouting hole at the current moment and the geometric interference area with adjacent holes. The degree of interference between holes is introduced into the optimization objective. With the dual objectives of maximizing the sealing efficiency of the corrected fracture section and the strength of the time-cured structure, the NSGA-II evolution algorithm is used to solve for the optimal combination of grouting pressure and grout dynamic viscosity in the next control cycle. During the optimization process, the upper and lower bounds of the variation space are dynamically adjusted according to the spatial interference penalty factor. While ensuring population diversity, the dangerous parameter combination of high pressure and low viscosity is actively suppressed. The above technical means enable each grouting hole to adaptively adjust the grouting pressure and grout viscosity according to the actual formation response, effectively avoiding construction risks such as cross-grouting, secondary splitting, and grouting blind zones caused by excessive compression of the grout diffusion front. Attached Figure Description

[0015] Figure 1 A flowchart illustrating the steps of a collaborative control method for grouting operations in building engineering based on edge computing, provided in one embodiment of the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Example 1, referring to Figure 1 This paper provides a collaborative control method for grouting operations in building engineering based on edge computing, which includes the following steps: Step S1: Collect construction data, grouting parameters, and engineering parameters; Step S2: Based on construction data, grouting parameters, and engineering parameters, execute a state estimation algorithm for each grouting hole to obtain an estimated value of the diffusion distance of the grout diffusion front at the current moment; Step S3: Based on the diffusion distance estimate and construction data, perform spatial geometric subdivision and interference quantization to calculate the control area and geometric interference area of ​​each grouting hole; Step S4: Based on the estimated values ​​of control area, geometric interference area, and diffusion distance, a multi-objective optimization algorithm is used to solve for the optimal combination of grouting parameters for the next control cycle. Step S5: Perform a safety verification on the optimal grouting parameter combination, and send out the parameters that pass the verification for execution.

[0018] This invention collects construction data, grouting parameters, and engineering parameters through edge computing devices, and sequentially performs state estimation, spatial geometric partitioning and interference quantification, multi-objective optimization solution, and safety verification to form a complete closed-loop collaborative control process. This invention introduces the estimation of the underground diffusion front of the grout, the quantification of the control area and interference area between boreholes into the grouting decision-making process, so that each grouting hole can dynamically adjust its own grouting parameters according to the actual response of the stratum. This effectively avoids cross-grouting, secondary stratum splitting, or grouting blind spots caused by the mutual compression of the grout diffusion front between boreholes, and improves the overall collaborative operation efficiency and construction safety of group borehole grouting.

[0019] In step S1, the construction data includes the grouting pressure and grouting flow rate during the grouting process; Grouting parameters include the set value of grout dynamic viscosity; Engineering parameters include: permeability of the soil in the grouting area, effective grouting layer thickness through which the grouting hole passes, environmental pore water pressure, effective porosity of the medium in the grouting area, crack initiation toughness constant of the grouting medium, radius of the grouting hole, and design target diffusion radius.

[0020] In one specific embodiment of the present invention, the grouting holes laid out in the target building project are numbered, and a construction coordinate system is constructed based on the construction plane to obtain the construction coordinates of all grouting holes. Edge computing devices are deployed at the construction site. These devices are connected to pressure sensors, flow meters, and programmable logic controllers (PLCs) of each grouting hole via a fieldbus. Each grouting unit is equipped with a dual-liquid dynamic mixing system controlled by a PLC. The dynamic viscosity of the injected grout can be changed in real time by adjusting the mixing volume ratio of liquids A and B. The edge computing devices perform data processing and construction decisions in a fixed control cycle. In this embodiment, the control cycle is preset to 30 seconds. In this embodiment, the relationship between the mixing volume ratio of liquids A and B and the final dynamic viscosity of the grout has been pre-calibrated through indoor tests before the formal grouting operation. This calibration relationship is stored in the edge computing device in the form of a viscosity lookup table. When the edge computing device determines the optimal viscosity value for the next control cycle through the optimization algorithm, it will synchronously query the lookup table, obtain the corresponding A and B liquid ratio value through linear difference, and send the ratio value and the optimal grouting pressure value to the programmable logic controller of the corresponding grouting unit. Within each control cycle, with a sampling interval of 1 second, the grouting pressure and grouting flow rate of all grouting holes are collected. For grouting hole i, the instantaneous grouting pressure collected at time t is denoted as... The unit is megapascal; the original value of the grouting flow rate is denoted as... The unit is liters per minute; Permeability of soil in grouting area The effective grouting layer thickness is derived from the engineering survey report. Derived from engineering design drawings, grouting hole radius Determined by drill bit specifications, ambient pore water pressure The effective porosity of the medium in the grouting area was obtained in real time by a vibrating wire piezometer embedded in the monitoring boreholes adjacent to the grouting area. The crack initiation toughness constant of the grouting medium was determined by on-site sampling and laboratory testing. Target diffusion radius, derived from engineering material test report. This requirement for the diffusion range of grout in a single hole is derived from the grouting operation design documents of building engineering. The comprehensive flow coefficient is calculated from the grouting hole during the test phase based on the stable injection flow rate and injection pressure, and by taking the logarithmic term when the grout spreads to a typical radius of 1.0 meter. It is used to comprehensively characterize the average flow capacity of the medium in the grouting area of ​​this project. The process of obtaining the comprehensive flow coefficient specifically includes: Before the formal grouting operation begins, a grouting hole is selected, clean water is used as the injection medium, and the grouting pump is set to run at a constant speed. From the moment the injection begins, the grouting pressure and grouting flow rate are continuously recorded at 1-second intervals. When the fluctuation range of the grouting flow rate at 10 consecutive sampling moments does not exceed ±5% of the average value of the 10 flow rates, the injection is determined to have entered a stable state. The grouting pressure at 10 sampling times was averaged. Take the average value of the grouting flow rate. In this embodiment, after the pressurized water reaches a stable state, the measured value is... , ,Right now ; Slurry diffusion radius The 1.0-meter radius is used because the design spacing of the grouting holes in this project is 2.0 meters. The grouting operation requires that the grout diffusion range of adjacent grouting holes overlap at the centerline between the holes. Therefore, the design target diffusion radius of the grout in a single hole is determined by this radius. It is half of 2.0 meters, i.e., 1.0 meter. The purpose of the water pressure test is to calibrate the comprehensive flow capacity of the medium when the slurry diffuses to this typical design radius, hence the value is taken as... Meters. Grouting hole radius Meters, calculate the value of the logarithmic term at this time. ; The radial Darcy flow formula, under the quasi-steady-state seepage assumption, takes the following form: ; in, This refers to the grouting flow rate, in m³ / s. The permeability of the soil in the grouting area, in m². The effective grout layer thickness traversed by the grouting hole, in meters (m). The grouting pressure is expressed in Pa. Environmental pore water pressure, in Pa. The dynamic viscosity of the injected fluid, in Pa·s. The current diffusion radius of the slurry, in meters. The radius of the grouting hole is in meters. The comprehensive flow coefficient is constructed, and its mathematical expression is as follows: ; The unit is m³, which characterizes the average flow capacity of the medium in the grouting area of ​​this project. It is used as a constant in subsequent control cycles and is introduced into... Then, the radial Darcy flow formula simplifies to: Substitute the data under stable injection conditions into the radial Darcy flow formula to obtain the overall flow coefficient; Substituting the steady-state data from the water pressure test, we obtain: ; in, To determine the dynamic viscosity of water at room temperature, take... ; For environmental pore water pressure, take multiplied by Convert MPa to Pa; In this embodiment, water pressure tests were conducted on three representative grouting holes. After calculating the three comprehensive flow coefficient values ​​according to the above steps, the arithmetic mean was taken to obtain the final comprehensive flow coefficient. The dynamic viscosity parameter of the grouting fluid is taken as the actual set viscosity value sent to grouting hole i in the previous control cycle. The initial viscosity ratio is determined at the beginning of the first control cycle of grouting. In this embodiment, an effective flow rate threshold is preset. ; The process of determining the effective flow rate threshold includes: before the formal grouting operation begins, closing all grouting valves at all grouting orifices, keeping the grouting pipeline in the same connection state as the formal grouting, starting the pressure sensor and flow meter to collect data, continuously recording the background noise reading of the flow meter in the zero flow state for 30 seconds, recording the instantaneous flow value once every 1 second, and obtaining a total of 30 sample data, as shown in Table 1; Table 1: ; Statistical analysis was performed on the 30 sample data in Table 1. The maximum value was 0.06 L / min, the average value was 0.033 L / min, and the standard deviation was 0.011 L / min. Considering that the noise amplitude of the flow meter may occasionally be slightly higher than the maximum value of this static test due to factors such as temperature drift and pipeline micro-vibration during long-term operation, a safety margin was reserved. 1.5 times the maximum value of the static test was taken and rounded up to 0.1 L / min as the effective flow rate threshold. Right now When the raw traffic is collected If the flow rate is below the effective flow rate threshold, it is considered an invalid injection; when the raw flow rate is collected... When the flow rate is greater than or equal to the effective flow rate threshold, it is determined to be an effective slurry injection.

[0021] In step S2, the state estimation algorithm is the Kalman filter algorithm, and the execution process includes: The radial distance and radial diffusion velocity of the grout in each grouting hole are constructed as a state vector; At each sampling time, the prior predicted state vector and prediction error covariance matrix at the current time are calculated using the state transition equation based on the state vector at the previous time. Perform observation validity assessment. The logic for observation validity assessment includes: If the grouting pressure at the sampling moment is greater than the environmental pore water pressure, and the grouting flow rate is greater than or equal to the effective flow velocity threshold, then the grouting flow rate is converted into volumetric flow rate, substituted into the radial Darcy flow formula inverse solution to obtain the observed diffusion radius as the observed value, and the observation matrix is ​​set; the Kalman gain matrix is ​​calculated; based on the Kalman gain matrix, the prior predicted state vector, the observed value, and the observation matrix, the posterior state estimation vector at the current moment is calculated, and the posterior estimation error covariance matrix is ​​updated; If the grouting pressure at the sampling time is less than or equal to the environmental pore water pressure, or the grouting flow rate is less than the effective flow velocity threshold, then the calculation of the observed value and Kalman gain is skipped, the radial diffusion velocity at the current time is forcibly set to 0, the radial distance is kept as the posterior state estimate of the previous time, and the posterior estimation error covariance matrix is ​​set to be equal to the posterior estimation error covariance matrix of the previous time plus the process noise covariance matrix; finally, the diffusion distance estimate of each grouting hole at each sampling time is output.

[0022] In one specific embodiment of the present invention, since the actual diffusion boundary of the grout in the building foundation medium is not visible, this embodiment executes a Kalman filter algorithm for each grouting hole to estimate the radial distance and velocity of the grout diffusion front at the current moment based on the state and physical motion law of the grout at the previous moment. The radial distance and velocity of the grouting hole are then used to construct a state vector, and the state vector of grouting hole i is: ,in, This represents the radial distance of slurry diffusion at the current moment, in meters. Radial diffusion velocity, in meters per second; The process of executing the Kalman filter algorithm to obtain the current estimated state vector includes: Set the initial conditions for the Kalman filter algorithm. The initial conditions specifically include: At the start of grouting The grouting holes have not yet received a large amount of grout; the initial radial distance is set. Meters correspond to the wetting range near the grouting hole wall, and the initial diffusion rate is set accordingly. meters per second; Define the posterior estimation error covariance matrix The process noise covariance matrix is ​​a 2×2 diagonal matrix with all diagonal elements being 0.01. It is a 2×2 diagonal matrix with all diagonal elements being 0.0001, used to describe the state disturbance caused by the non-uniformity inside the building; Observation noise variance The value is set to 0.0025 m² to reflect the uncertainties caused by sensor errors and model simplification. After acquiring the collected data at time t, a priori state prediction is performed using a state transition model. The state transition adopts a uniform motion model. Considering that the acquisition interval is only 1 second and the pressure and flow rate changes relatively slowly, it can be approximated as quasi-steady-state radial seepage. The input to the state transition model is the previous time step, i.e. State vector at time step ; The mathematical expression of the state transition equation constructed in this embodiment is: ; in, Grouting holes The predicted state vector at time t, Grouting holes At any moment The state vector, This is the state transition matrix; The data acquisition interval is 1 second in this embodiment; In this embodiment, the state transition matrix is: The state transition equation states that the estimated diffusion distance at the current moment is equal to the distance at the previous moment plus the displacement generated by the velocity within 1 second, while assuming that the velocity remains constant in short time intervals. The predicted state vector at the current time step is obtained using the state transition equation, and the prediction error covariance matrix is ​​calculated. The mathematical expression for the prediction error covariance matrix is ​​as follows: ; in, Let be the prediction error covariance matrix at time t. Here is the state transition matrix. This is the transpose of the state transition matrix. for The posterior estimation error covariance matrix at time 1. The process noise covariance matrix; After completing the prior prediction, the observation validity judgment and posterior update are performed. The judgment logic for the observation validity judgment includes: like Greater than ,and Greater than or equal to This indicates that effective grout injection is underway at this point. Multiply by conversion factor This is converted into a flow rate in cubic meters per second. Then, substitute the radial Darcy flow formula into the inverse solution to obtain the observed diffusion radius, and use the observed diffusion radius as the observed value. The mathematical expression for the observed value is: ; in, Let i be the observed diffusion radius of the grouting hole at time t. Where is the radius of the grouting hole. For the comprehensive flow coefficient, This is the actual set value of the dynamic viscosity of the grout in grouting hole i at time t. This is the converted instantaneous grouting flow rate; and Multiply by MPa Convert to Then, the dimensions are matched with permeability and viscosity to prevent reverse calculations. If the value is too small, it will cause non-positive logarithms in subsequent calculations, so the calculation result is truncated to protect it. like Then force order The 0.001 meter is the preset limit for the thickness compensation of the attached manifold, and the observation matrix is ​​set. That is, only the radial distance is observed directly, and the velocity is not observed; The Kalman gain is calculated using the calculated prediction error covariance matrix. The mathematical expression for the Kalman gain is: ; in, Let be the Kalman gain matrix at time t. Let be the prediction error covariance matrix at time t. For the observation matrix, To observe the noise variance; The state is corrected posteriorly to obtain the posterior state estimate vector at time t. The mathematical expression for the posterior state estimate vector is: ; in, Let be the posterior state estimate vector at time t. Let be the prior predicted state vector at time t. Let be the Kalman gain matrix at time t. Let be the observed diffusion radius at time t. For the observation matrix, Let be the prior predicted state vector at time t; Update the posterior estimation error covariance matrix. The mathematical expression for the posterior estimation error covariance matrix is ​​as follows: ; in, Let be the posterior estimation error covariance matrix at time t. Let be the Kalman gain matrix at time t. It is a 2×2 identity matrix. For the observation matrix, Let be the prediction error covariance matrix at time t; like Less than or equal to ,or Less than This indicates that there was no effective slurry injection at that moment. Considering that the slurry is a non-Newtonian fluid with yield stress, it will quickly stop diffusing after losing effective injection pressure due to the viscous resistance of the medium pores. If the uniform motion prediction model of the state transition equation of Kalman filter is continued to be used, the diffusion distance will cause an infinite erroneous divergence that deviates from the physical reality. Skip the above calculation of observations and gains, do not use the prior prediction, and instead force the introduction of physical flow interruption constraints; Let the diffusion speed at the current moment be... The radial distance remains the value from the previous time step, meaning the posterior state estimate is directly assigned to the value from the previous time step. Let the posterior estimation error covariance matrix be... Equal to the error covariance matrix of the previous time step Add process noise This is to reasonably characterize the natural accumulation of uncertainty in the pump's state during the pump shutdown period; The final output is the estimated slurry diffusion distance for each well at each sampling time. .

[0023] This invention applies the Kalman filter algorithm to the real-time estimation of the grout diffusion front in grouting holes. It also incorporates physical flow interruption constraints to address potential invalid injection conditions during grouting, such as grouting pressure below pore water pressure or flow rate below a threshold. This forces the diffusion velocity to zero, effectively preventing distance divergence errors in the state prediction model caused by the assumption of uniform motion during pump shutdown or pressure loss. Furthermore, it maintains the physical rationality of the state estimation even under non-ideal conditions such as grouting intervals, pipeline blockage, or pressure fluctuations, providing continuous and reliable diffusion distance input for subsequent spatial geometric partitioning and optimization decisions.

[0024] Step S3, the process of spatial geometric subdivision and interferometric quantization, specifically includes: At the beginning of each control cycle, obtain the estimated values ​​of grouting pressure and diffusion distance from the previous sampling time. Based on the grouting pressure of each grouting hole, a weighting factor is calculated. The process of calculating the weighting factor specifically includes: Calculate the maximum and minimum pressure values ​​among all grouting holes, calculate the normalized value for each grouting hole, and set the weighting factor to the normalized value plus 1; Using the planar coordinates of each grouting hole as seed points and weighting factors as weights, the multiplicative weighted Voronoi diagram algorithm is executed to generate the weighted Thiessen polygon region corresponding to each grouting hole; For any point on a two-dimensional plane, divide the actual geometric distance from the point to the grouting hole by the weighting factor of the grouting hole to obtain the weighted distance. Divide each point into the grouting hole that minimizes its weighted distance to obtain the weighted Thiessen polygon region corresponding to each grouting hole. Based on the ordered vertex coordinate sequence of the weighted Thiessen polygon region corresponding to each grouting hole, the area enclosed by the weighted Thiessen polygon region is calculated using the shoelace formula, which serves as the control area of ​​the grouting hole.

[0025] In one specific embodiment of the present invention, at the beginning of each control cycle, the grouting pressure collected in real time at the previous sampling time, i.e., the last second of the previous control cycle, is acquired. Based on grouting pressure Performing spatial geometric partitioning and interferometric quantization, the specific process includes: Pressure on each hole Perform real-time dimensionless transformation to generate weighting factors. Let the maximum pressure among all grouting holes at the start of the current control cycle be denoted as . The minimum value is Calculate the normalized value for each grouting hole. , and then From this, we obtain The value ranges from 1 to 2, and it is a dimensionless weighting coefficient. The higher the pressure of the grouting hole, the greater its weighting coefficient. The closer the pressure is to 2, the lower the pressure of the hole. The closer to 1; If the pressure in all grouting holes is consistent, then all All are set to 1.5; Using the plane coordinates of each grouting hole As seed point, with As weighting factors, the multiplicative weighted Voronoi diagram algorithm from the computational geometry library CGAL is used to generate weighted Thiessen polygons. The input to the multiplicative weighted Voronoi diagram algorithm is the two-dimensional coordinates of each grouting hole and its weighting factors. The output is a weighted Thiessen polygon region corresponding to each grouting hole. The weighted Thiessen polygon region is represented by an ordered sequence of vertex coordinates. For any point p(x,y) on a two-dimensional plane, the mathematical expression for its weighted distance to the grouting hole i is: ; in, Let p(x,y) be the weighted distance from point p(x,y) to grouting hole i, where (x,y) are the coordinates of any point p in the plane. Let i be the construction coordinate of the grouting hole. , is the weighting factor for grouting hole i; In the mathematical expression for weighted distance, the numerator is the actual geometric distance in meters, and the denominator is... Since it is a dimensionless number, the weighted distance The unit is still meters, which can be directly used for geometric region division, dividing point p to make... The smallest grouting hole is obtained, thus yielding the weighted Thiessen polygon region corresponding to each grouting hole; Since the region boundary generated by the multiplicative weighted Voronoi diagram is a circular arc, the computational geometry library can discretize it into polygons with a preset precision. The discretized polygons are the weighted Thiessen polygon regions corresponding to each grouting hole. Based on the ordered vertex coordinate sequence of the weighted Thiessen polygon region corresponding to the grouting hole, the area enclosed by the closed plane region is calculated using the shoelace formula, and this area is used as the control area. ; The higher the pressure of the grouting hole, the more... The larger the value, the smaller the weighted distance, thus occupying a larger area in the geometric partition, and therefore controlling the area. This reflects the actual control range of the grout in grouting hole i within the building foundation medium at the current moment.

[0026] This invention dynamically generates weighted Thiessen polygons based on the real-time grouting pressure of each grouting hole, and divides the control area of ​​each hole by weighted distance. The grouting hole with higher pressure gets a larger control area, which corresponds to the physical phenomenon that the grout is more likely to expand outward under high pressure. This allows the control area to reflect the actual influence range of each grouting hole in the formation at the current moment. Compared with the method of using a fixed radius circle or an equal weighted Voronoi diagram, this invention can dynamically adjust the control boundary by capturing the changes in grouting pressure.

[0027] Step S3, the process of calculating the geometric interference area, specifically includes: In the interference calculation of the buffer warning zone, the weighted Thiessen polygon region of each grouting hole is shifted outward by a preset safe buffer distance to generate a buffer extended polygon. For all adjacent grouting hole pairs, calculate the area of ​​the intersection region of the two buffer extended polygons, where adjacent grouting hole pairs are grouting hole pairs with a common edge in the corresponding weighted Thiessen polygon regions; For each grouting hole, the geometric interference area of ​​the intersection region generated by it and all adjacent grouting holes is summed.

[0028] In a specific embodiment of the present invention, buffer warning zone interference calculation is performed on the weighted Thiessen polygon region, and the specific process includes: Preset safety buffer distance Meters, this value was obtained through in-situ splitting tests at the engineering site. The specific procedures of the test included: Before the formal grouting operation begins, a location is selected at the edge of the grouting area, and a central test hole and four monitoring holes are drilled. The central test hole is used to apply grouting pressure and induce fracturing. The four monitoring holes are arranged at different distances from the central test hole, with hole spacing of 0.3 meters, 0.5 meters, 0.8 meters and 1.2 meters respectively. The diameter, depth and drilling process of all holes are consistent with those of the formal grouting holes. Clean water was injected into the central test hole in stages, starting with a grouting pressure of 0.5 MPa and increasing by 0.5 MPa at each stage. Each pressure level was maintained for 120 seconds before moving to the next stage. During each pressure maintenance period, pressure sensors and flow meters installed at the orifices of four monitoring holes were used to continuously record whether pressure response and water discharge occurred at each monitoring hole. If any of the following phenomena first appeared at a monitoring hole, it was determined that a splitting connection had occurred between that monitoring hole and the central test hole at that pressure level. The specific phenomena included: The pressure sensor reading at the orifice of the monitoring hole increased by more than 0.1 MPa from the ambient value within 30 seconds; Continuous water flow was observed from the monitoring well; During the step-by-step pressurization process, the grouting pressure value corresponding to the first occurrence of splitting and penetration phenomenon in each monitoring hole was recorded. The test results are shown in Table 2. Table 2: ; Based on the above test results, the limit safety distance to avoid splitting is determined to be 0.5 meters. In the collaborative grouting operation of multiple holes, if the distance between the grout diffusion fronts of adjacent grouting holes is less than 0.5 meters, there is a real risk of secondary splitting of the grouting medium. Therefore, a safety buffer distance of 0.5 meters is taken. In the interference calculation of the buffer warning zone, the polygonal boundary of each grouting hole is shifted outward by 0.5 meters as the buffer expansion boundary. When the buffer expansion area overlaps, it means that the distance between adjacent holes has entered the splitting danger range. The weighted Thiessen polygon region of each grouting hole is shifted outward in the normal direction. Distance, generating buffer extended polygons, and calculating the intersection area of ​​buffer extended polygons can be accomplished by calling the buffer analysis and face overlay analysis functions in the computational geometry library; The weighted Thiessen polygon region of each grouting hole is shifted outward in the normal direction. The specific process of distance is as follows: For each edge of the weighted Thiessen polygon region, translate along its outward normal direction. By extending adjacent translation edges to their intersection, new vertices are obtained, thus constructing an extended polygon, which is represented by an ordered new sequence of vertex coordinates; For all adjacent grouting hole pairs Calculate the area of ​​the intersection region of two extended polygons, where adjacent grouting hole pairs refer to grouting hole pairs formed by grouting holes with a common edge in the corresponding weighted Thiessen polygon regions; For grouting hole i, the geometric interference area corresponding to grouting hole i is obtained by summing the intersection areas of grouting hole i with all adjacent grouting holes. The unit is square meters. If the extended polygons of an adjacent pair of grouting holes do not overlap, then the corresponding intersection area is 0. The larger the value, the more intense the compression between the diffusion front of grouting hole i and adjacent holes, and the higher the risk of grout cross-contamination or splitting.

[0029] This invention generates a buffer extension polygon by shifting the weighted Thiessen polygon of each grouting hole outwards by a preset safety buffer distance in the normal direction. The intersection area of ​​the extension polygons between adjacent holes is calculated as the geometric interference area. The safety buffer distance is derived from field splitting tests and can quantitatively characterize the minimum allowable safe distance between the grout fronts of adjacent grouting holes. The size of the geometric interference area directly reflects the severity of the current risk of inter-hole compression and cross-contamination, thus transforming the group hole interference from a qualitative judgment into a quantifiable optimization index, providing a clear input for the calculation of penalty factors and the compression of the mutation space in multi-objective optimization algorithms.

[0030] In step S4, the parameter combination optimized by the multi-objective optimization algorithm is grouting pressure and grout dynamic viscosity; The optimization objectives include: maximizing the crack sealing efficiency and maximizing the strength of the age-cured structure; The process of calculating the maximum crack sealing efficiency specifically includes: Predicted grout diffusion distance after one control cycle under candidate grouting pressure and grout dynamic viscosity. The mathematical expression is: ; in, To predict the slurry diffusion distance, This is the estimated slurry diffusion distance output from step one at the current moment. For the comprehensive flow coefficient, For candidate grouting pressure, To control the cycle duration, The effective grout layer thickness through which the grouting hole penetrates. The effective porosity of the medium in the grouting area. The dynamic viscosity of the candidate slurry; The sealing effectiveness of the fracture section is calculated based on the ratio of the predicted grout diffusion distance to the designed target diffusion radius. The mathematical expression for the sealing effectiveness of the fracture section is as follows: ; in, To improve the sealing effectiveness of the fracture cross section, To predict the slurry diffusion distance, To design the target diffusion radius; when Reaching or exceeding hour, This indicates complete closure; candidate pressure P and viscosity. Direct impact Thus change This ensures the effective search for blocking targets. A penalty factor based on spatial geometric interference is introduced to modify the crack sealing effectiveness. The modification process specifically includes: The ratio of the geometric interference area to the control area of ​​each grouting hole is used as the spatial penalty factor, and the spatial penalty factor is truncated. The crack sealing efficiency is corrected based on the truncated spatial penalty factor. The mathematical expression for the corrected crack sealing efficiency is as follows: ; in, To improve the sealing effectiveness of the corrected fracture cross section, The sealing effectiveness of the crack section before correction. The truncated spatial penalty factor. For candidate grouting pressure, This is the upper limit for the grouting pressure search.

[0031] In one specific embodiment of the present invention, the control area is extracted. Geometric interference area and diffusion distance estimate Solve for the optimal combination of grouting parameters for the next control cycle of grouting hole i. The variables to be optimized in the grouting parameter combination are the grouting pressure P and the dynamic viscosity of the grout in the next cycle. The control cycle duration is fixed at 30 seconds; The optimization process employs a multi-objective evolutionary algorithm based on the NSGA-II framework, but the mutation operation and fitness evaluation are subject to forced intervention by the spatial geometric state. The optimization objectives during the optimization process include maximizing the sealing efficiency of the crack cross section and maximizing the strength of the age-cured structure; The process of calculating the sealing effectiveness of a fracture cross section specifically includes: Predicting in candidate parameters Under the action, the slurry diffusion distance after one control cycle ; Predicting slurry diffusion distance The mathematical expression is derived based on the radial diffusion volume equilibrium relationship, and the mathematical expression for predicting the injection volume is: ; in, To predict the injection volume, For the comprehensive flow coefficient, For candidate grouting pressure, The dynamic viscosity of the candidate slurry. To control cycle duration; Meanwhile, the volume relationship of the cylindrical diffusion structure satisfies the mathematical expression: ; in, To predict the injection volume, The effective grout layer thickness through which the grouting hole penetrates. The effective porosity of the medium in the grouting area. To predict the slurry diffusion distance, This is an estimate of the diffusion distance; Solving the two expressions simultaneously yields the mathematical expression for predicting the slurry diffusion distance. It should be further noted that this mathematical expression approximates the average injection flow rate within the current control cycle using the combined flow coefficient, neglecting the influence of the slowly varying logarithmic term in the radial Darcy formula. The resulting prediction bias will be corrected by re-estimating the actual diffusion distance using a Kalman filter in each control cycle. Closed-loop correction is used to ensure the accuracy and convergence of the entire control process; A penalty factor based on spatial geometric interference is introduced to modify the crack sealing effectiveness. The specific process includes: For grouting hole i, the geometric interference area and control area The ratio result is used as a spatial penalty factor. If the control area of ​​grouting hole i If it is 0, then the space penalty factor is zero. ; To prevent spatial penalty factors because and If the area overlap exceeds 1, resulting in a negative modified performance, the spatial penalty factor is truncated with an upper limit of 1. The penalty factor reflects the severity of spatial interference between the current grouting hole i and its neighboring holes within the buffer zone; The effectiveness of crack sealing is modified based on a penalty factor. The significance of correcting the upper limit of the grouting pressure set is to address spatial interference. In severe cases, the higher the grouting pressure P in the candidate parameter combination, the heavier the penalty it suffers and the lower the sealing efficiency value. As a result, in the evolutionary selection, high-pressure dangerous parameters will be quickly eliminated due to their extremely poor fitness, while low-pressure parameters can retain some fitness and block parameter combinations that cause cross-grouting and splitting.

[0032] This invention aims to maximize both the sealing efficiency of fracture sections and the strength of age-cured structures. It introduces a spatial penalty factor based on the ratio of geometric interference area to control area into the correction calculation of sealing efficiency. When the geometric interference area is large, high-pressure grouting will suffer severe penalties, causing the evolutionary algorithm to automatically eliminate dangerous parameter combinations of high pressure and low viscosity, while low-pressure parameters can still retain some fitness. Thus, while pursuing sealing effect, it actively suppresses parameters that induce formation fracturing, achieving joint optimization of grouting pressure, viscosity and inter-hole interference risk.

[0033] The mathematical expression for the strength of age-cured structures is: ; in, For age-cured structural strength, The first fitting coefficient, The dynamic viscosity of the candidate slurry. For reference viscosity, is the second fitting coefficient.

[0034] The process of calculating the strength of age-cured structures specifically includes: The strength F2 of the age-cured structure is calculated using an empirical formula fitted from indoor grout-bonded mass strength tests. The process of determining the empirical formula includes: The grout used in this project is a mixed grouting material of A and B. Its dynamic viscosity μ can be continuously changed in the range of 0.05 Pa·s to 0.5 Pa·s by adjusting the mixing volume ratio of A and B. In order to establish a quantitative relationship between grout viscosity and the strength of the solidified stone body, a grout stone body strength test was carried out in the laboratory. Five different mixing ratios were selected to achieve slurry dynamic viscosities μ of 0.05 Pa·s, 0.10 Pa·s, 0.20 Pa·s, 0.35 Pa·s, and 0.50 Pa·s, respectively. For each ratio, three cylindrical specimens with a diameter of 50 mm and a height of 100 mm were prepared according to the standard test procedure. After curing for 28 days under standard curing conditions (temperature 20 ± 2℃, relative humidity ≥ 95%), uniaxial compressive strength tests were conducted on a universal testing machine at a loading rate of 0.5 MPa / s. The arithmetic mean of the compressive strength of the three specimens under each ratio was taken as the age-cured structural strength corresponding to that viscosity. The test results are shown in Table 3.

[0035] Table 3: ; Take the natural logarithm of the viscosity μ in Table 3. As the independent variable, among which, The reference viscosity is introduced to make the independent variable of the logarithmic operation dimensionless. The corresponding average intensity F2 is used as the dependent variable, and a scatter plot is drawn. The strength of cementitious aggregates primarily originates from the gel structure formed by the hydration reaction of active components in the cement grout. Higher viscosity indicates a higher concentration of active components and a lower water-cement ratio, resulting in a denser gel and higher aggregate strength. According to the strength theory of cement-based materials, strength and gel volume fraction approximately satisfy a logarithmic relationship, which can be expressed as a linear relationship. As a mathematical structure of empirical formulas, in which, The first fitting coefficient, The second fitting coefficient; The 5 groups in Table 3 Substitute the data into the least squares linear regression method and find the set of errors that minimizes the sum of squared errors. Thus, the mathematical expression for the strength of the age-cured structure in this embodiment is obtained.

[0036] By using the logarithmic empirical formula between the strength of the time-cured structure and the dynamic viscosity of the grout, the strength of the stone body corresponding to any candidate viscosity can be quickly evaluated. This allows multi-objective optimization to quantitatively balance the sealing performance and curing strength, avoiding the problem of excessively thin grout and insufficient curing strength caused by simply pursuing a large diffusion radius. Thus, it ensures the grouting coverage while taking into account the mechanical properties of the reinforced strata.

[0037] Step S4, the evolution process of the multi-objective optimization algorithm, specifically includes: The population size and number of iterations of the evolutionary algorithm are preset. Samples are selected from the parent population through tournament selection. Simulated binary crossover is performed on the samples with a set probability to generate offspring. Mutation operation is performed on all generated offspring individuals, where the mutation operation is determined by the current spatial topology. The process of performing mutation operations specifically includes: Extract the spatial penalty factor value after the grouting hole is truncated and compare it with the preset penalty threshold; if the truncated spatial penalty factor value is less than the penalty threshold, it is determined that the current grouting hole is in a safe and cooperative state, and regular Gaussian mutation is performed. If the truncated space penalty factor value is greater than or equal to the penalty threshold, the current grouting hole is determined to be in a collision danger state. The directional dynamic compression of the variation space is performed, the upper limit of the allowable variation of the candidate grouting pressure is temporarily lowered, the lower limit of the allowable variation of the candidate grout dynamic viscosity is temporarily raised, and conventional Gaussian variation is performed within the compressed range. After the optimization algorithm completes its iterations, it outputs the Pareto front non-dominated solution set on the optimization objective. All non-dominated solutions on the Pareto front are sorted, and the solution ranked first is extracted as the candidate optimal parameter combination. The evaluation algorithm is called to calculate the stress intensity factor corresponding to the parameter combination. If the stress intensity factor is less than the crack initiation toughness constant of the grouting medium, the parameter combination is taken as the final safe and feasible solution. If the stress intensity factor is greater than or equal to the crack initiation toughness constant of the grouting medium, then the next non-dominated solution is extracted sequentially for verification until a parameter combination that satisfies the stress intensity factor being less than the crack initiation toughness constant of the grouting medium is found. The final grouting pressure and grout dynamic viscosity values ​​are sent to the corresponding grouting unit's programmable logic controller for execution.

[0038] In a specific embodiment of the present invention, the two optimization objectives of maximizing the sealing efficiency of the modified crack section and maximizing the strength of the age-cured structure are transformed into the problem of minimizing their opposites, which is used to uniformly apply the non-dominated sorting mechanism of the NSGA-II evolution algorithm. In this embodiment, the population size of the evolutionary algorithm is set to 100, the number of iterations is set to 50, the chromosome is encoded by the real number (P,μ), the search range of P is [0.5,5]MPa, and the search range of μ is [0.05,0.5]Pa·s; The initial population is generated uniformly and randomly within the above range. In each generation, a tournament selection method with a tournament size of 2 is used, that is, two individuals are randomly selected each time, and individuals with higher fitness or higher non-dominant ranking are selected from the parent population. For each sample, simulated binary crossover with a distribution exponent of η_c is performed with a probability of 0.9 to generate offspring, where η_c = 20. For samples with a probability of 0.1 where no crossover occurs, the sample is directly copied into the offspring. Perform a specific mutation operation on all generated offspring individuals. The mutation operation is determined by the current spatial topology. The specific process of the mutation operation includes: Extract the current spatial penalty factor λ_i' value of grouting hole i and compare it with the preset penalty threshold. In comparison, in this embodiment ; The penalty threshold was determined from the data of the engineering test section. Specifically, in the test section before formal construction, three adjacent grouting holes of grouting hole i were selected, and 10 sets of continuous grouting tests were carried out for each candidate penalty threshold. The overall sealing rate and the number of secondary splitting events were recorded, and the average values ​​are shown in Table 4. Table 4: ; As shown in Table 4, when When λ = 0.25, the average sealing rate reached 93.6%, and no secondary splitting occurred. When λ0 increased to 0.30, although the sealing rate slightly improved (94.8%), two splitting incidents occurred. Considering both sealing efficiency and construction safety, the selected... =0.25 is used as the optimal penalty threshold; During the mutation operation, if If the current grouting hole is determined to be in a safe and coordinated state, a regular Gaussian mutation is performed, the specific process of which includes: For pressure P, with a mutation probability of 0.9, a normally distributed random perturbation with a mean of 0 and a standard deviation of 5% of the current gene value is superimposed on its current gene value. Similarly, for viscosity μ, a normally distributed perturbation with a mean of 0 and a standard deviation of 5% of its current gene value is superimposed with a probability of 0.9. If the mutated parameter exceeds the preset safe search range, i.e. , If so, then cut it off to the nearest boundary; like The current grouting hole is determined to be in a collision hazard state. In order to maintain population diversity while guiding the optimization direction away from the hazard zone, a directional dynamic compression of the mutation space is forcibly executed. The specific process includes: The upper limit of allowable variation for the candidate pressure P is temporarily lowered to 85% of the original upper limit, meaning the upper limit of the compressed pressure is... The lower limit of allowable variation for candidate viscosity μ is temporarily raised to 110% of the original lower limit, meaning the lower limit of the compressible viscosity is... ; Within this range of safety parameters after contraction, i.e. , Continue with regular Gaussian mutation. If the parameters after the superimposed perturbation exceed the new boundary of the dynamic contraction, then forcibly truncate them to the nearest new boundary. The mutation process forcibly eliminates the dangerous combination of high pressure and low viscosity, thereby reducing geometric interference, avoiding continuous cumulative decline of the population, and achieving safe real-time closed-loop control.

[0039] After 50 generations of evolution, the algorithm output is optimized towards the target. and The Pareto front undominated solution set contains several undominated solutions. Each undominated solution represents a set of grouting parameters that cannot be surpassed by other solutions in maximizing the modified sealing effectiveness and the strength of the time-cured structure. Sort all nondominated solutions on the Pareto front. The sorting process specifically includes: According to the modified sealing effectiveness Sort the values ​​in descending order from largest to smallest. If there exists For solutions with the same value, the strength of the solidified structure shall prevail. The values ​​are sorted in descending order from largest to smallest, and solutions with higher strength are selected first. Extract the top-ranked solution as a candidate optimal parameter. ; Call the evaluation algorithm to calculate this set of parameters. The corresponding stress intensity factor, if Less than the crack initiation toughness constant of the grouting medium Then the parameter As the final safe and feasible solution; the space penalty factor in the optimization process. The aim is to suppress the risk of inter-hole interference caused by the mutual compression and cross-contamination of grout diffusion fronts between adjacent grouting holes. The evaluation algorithm is used to evaluate candidate combinations of grouting parameters. The risk of medium cracking caused by the medium surrounding the grouting hole itself; like If the crack initiation toughness constant of the grouting medium is greater than or equal to the value of the crack initiation toughness constant, then according to the order of all non-dominated solutions on the Pareto front, the next solution is extracted sequentially for verification until a parameter combination that satisfies the safety constraint is found. If no solution is found after traversing the entire set of non-dominated solutions, the condition is considered closed. The solution is less than the crack initiation toughness constant of the grouting medium; If no safe and feasible solution is found after traversing all solutions, it means that all non-dominated solutions on the Pareto front within the current control period have failed. The failure to pass the safety check occurs when the medium in the grouting area is close to the critical state of crack initiation and the stress intensity factor of all candidate parameter combinations is too high. To prevent the control system from falling into a dead loop due to repeated searches without a solution and thus being unable to issue commands, the underlying safety protection mechanism is triggered: it directly skips the optimization results of the evolution and forces the use of the preset most conservative parameter boundary value, namely the grouting pressure. slurry viscosity As the final safe and feasible solution, The selected grouting pressure and viscosity values ​​are sent as digital signals to the programmable logic controller (PLC) of the corresponding grouting unit. The PLC adjusts the opening of the proportional overflow valve accordingly and simultaneously adjusts the proportioning valve of the dual-liquid dynamic mixing system to match the set optimal viscosity μ_opt, and then performs the grouting operation for the next 30 seconds.

[0040] This invention introduces a spatial penalty factor into the mutation operation of the multi-objective evolution algorithm. When the geometric interference exceeds a set threshold, the upper bound of the grouting pressure mutation is directionally compressed and the lower bound of the viscosity mutation is increased, forcing the population to search within the safe region of low pressure and high viscosity. When all non-dominated solutions of the Pareto front fail to pass the stress intensity factor safety check, the bottom-level protection mechanism is triggered to directly adopt the most conservative parameters. This dynamic mutation and protection strategy ensures that a safe combination of grouting parameters can be output under any conditions, avoiding control interruption due to no optimization solution, and greatly enhancing the reliability and robustness of field applications.

[0041] The evaluation algorithm takes the candidate grouting pressure, the continuous grouting time in the current control cycle, and the dynamic viscosity of the grout as inputs, calculates the stress intensity factor at the crack tip, and takes the stress intensity factor as the output. The process of evaluating the algorithm specifically includes: The cumulative injection volume of the grout under the action of candidate parameters is calculated. The mathematical expression for the cumulative injection volume is: ; in, For the cumulative injection volume, For the comprehensive flow coefficient, For grouting pressure, For continuous grouting time, This is the dynamic viscosity of the slurry; P is in MPa, multiplied by... Convert to Pa, so that the unit of V is m³; The fracture half-length is calculated based on the cumulative injected volume. The mathematical expression for the fracture half-length is: ; in, For half the length of the crack, For the cumulative injection volume, The effective porosity of the medium in the grouting area; The stress intensity factor is calculated based on the grouting pressure and the half-length of the fracture. The mathematical expression for the stress intensity factor is as follows: ; in, Stress intensity factor For grouting pressure, The length of the fissure is half its length.

[0042] In one specific embodiment of the present invention, to prevent secondary fracturing of the building medium caused by grouting parameters, an evaluation algorithm based on fracture mechanics and grout seepage balance is pre-installed in the edge computing device. This evaluation algorithm uses the grouting pressure P and the continuous grouting time within the current control cycle as parameters. and slurry dynamic viscosity As input, output is the stress intensity factor at the crack tip. The process of calculating the stress intensity factor at the crack tip specifically includes: Calculation in parameters The cumulative injected volume of the grout is determined by the grout's entry into the building medium. The grout then fills and expands existing micro-cracks within the medium. In this embodiment, the cracks formed during grouting are approximated as coin-shaped radial cracks, assuming the cracks extend outwards from the grouting hole in a disk-like shape. The volume balance relationship indicates that the total volume of the injected grout is equal to the total volume of the crack space filled by the grout. For coin-shaped cracks, the total volume of the crack space is the volume of the disk multiplied by the proportion of the medium pores actually occupied by the grout. To simplify calculations and avoid introducing too many material parameters, the commonly used equivalent sphere assumption in engineering is adopted, assuming the injected grout fills a spherical pore space with radius *a*. The volume of the sphere is (4 / 3)πa³, where *a* is the half-length of the crack, i.e., the radius of the sphere. Since the crack space is not a complete vacuum, the proportion of the original pore volume displaced and filled by the grout is the effective porosity. Therefore, the volume balance relationship is that the total volume of the fracture space is (4 / 3)π. a³, based on the volume equilibrium relationship Solving for the solution yields the mathematical expression for the half-length of the crack.

[0043] A collaborative control system for grouting operations in building engineering based on edge computing includes a data acquisition module, an estimation module, a data segmentation module, an optimization module, and a verification module. The data acquisition module is used to collect construction data, grouting parameters, and engineering parameters. The estimation module is used to perform a state estimation algorithm on each grouting hole based on construction data, grouting parameters and engineering parameters, and obtain the estimated value of the diffusion distance of the grout diffusion front at the current moment; The subdivision module is used to perform spatial geometric subdivision and interference quantization based on construction data, and to calculate the control area and geometric interference area of ​​each grouting hole; The optimization module is used to solve the optimal combination of grouting parameters for the next control cycle based on the estimated control area, geometric interference area, and diffusion distance using a multi-objective optimization algorithm. The verification module is used to perform safety verification on the optimal grouting parameter combination and send out the parameters that pass the verification for execution.

[0044] This invention utilizes the Kalman filter algorithm to estimate the radial distance of the grout diffusion front at each grouting hole in real time, overcoming the technical problem of the difficulty in directly measuring the underground grout boundary. Based on this, a weighted Thiessen polygon is dynamically generated based on the grouting pressure to quantify the effective control area of ​​each grouting hole at the current moment and the geometric interference area with adjacent holes. The degree of inter-hole interference is introduced into the optimization objective. With maximizing the sealing efficiency of the corrected fracture section and the strength of the time-cured structure as dual objectives, the NSGA-II evolutionary algorithm is used to solve for the optimal combination of grouting pressure and grout dynamic viscosity in the next control cycle. During the optimization process, the variation is dynamically adjusted according to the spatial interference penalty factor. The upper and lower boundaries of the space, while ensuring population diversity, actively suppress dangerous parameter combinations of high pressure and low viscosity, and use fracture mechanics model to verify the stress intensity factor to ensure that all issued parameters are lower than the crack initiation toughness of the medium. The above technical means enable each grouting hole to adaptively adjust the grouting pressure and grout viscosity according to the actual response of the stratum, effectively avoiding construction risks such as cross-contamination, secondary splitting and grouting blind zone caused by excessive compression of the grout diffusion front. It significantly improves the overall sealing efficiency of group hole grouting and the uniformity of solidified body strength, while reducing manual intervention and trial and error costs, and provides a reliable and intelligent collaborative control scheme for large-scale grouting operations in construction engineering.

[0045] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0046] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A collaborative control method for grouting operations in building engineering based on edge computing, characterized in that, Includes the following steps: Step S1: Collect construction data, grouting parameters, and engineering parameters; Step S2: Based on the construction data, grouting parameters, and engineering parameters, perform a state estimation algorithm on each grouting hole to obtain an estimated value of the diffusion distance of the grout diffusion front at the current moment; Step S3: Based on the construction data, perform spatial geometric subdivision and interference quantization to calculate the control area and geometric interference area of ​​each grouting hole; Step S4: Based on the estimated values ​​of the control area, geometric interference area, and diffusion distance, a multi-objective optimization algorithm is used to solve for the optimal combination of grouting parameters for the next control cycle. Step S5: Perform a safety verification on the optimal grouting parameter combination, and send out the parameters that pass the verification for execution.

2. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 1, characterized in that, In step S1, the construction data includes the grouting pressure and grouting flow rate during the grouting process; The grouting parameters include the set value of the dynamic viscosity of the grout. The engineering parameters include: the permeability of the soil in the grouting area, the effective thickness of the grouting layer through which the grouting hole passes, the environmental pore water pressure, the effective porosity of the medium in the grouting area, the crack initiation toughness constant of the grouting medium, the radius of the grouting hole, and the design target diffusion radius.

3. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 2, characterized in that, In step S2, the state estimation algorithm is a Kalman filter algorithm, and the execution process includes: The radial distance and radial diffusion velocity of the grout in each grouting hole are constructed as a state vector; At each sampling time, the prior predicted state vector and prediction error covariance matrix at the current time are calculated using the state transition equation based on the state vector at the previous time. Perform an observation validity determination, the logic of which includes: If the grouting pressure at the sampling moment is greater than the environmental pore water pressure, and the grouting flow rate is greater than or equal to the effective flow velocity threshold, then the grouting flow rate is converted into volumetric flow rate, substituted into the radial Darcy flow formula to obtain the observed diffusion radius as the observed value, and the observation matrix is ​​set; the Kalman gain matrix is ​​calculated; based on the Kalman gain matrix, the prior predicted state vector, the observed value, and the observation matrix, the posterior state estimation vector at the current moment is calculated, and the posterior estimation error covariance matrix is ​​updated; If the grouting pressure at the sampling time is less than or equal to the environmental pore water pressure, or the grouting flow rate is less than the effective flow velocity threshold, then the calculation of the observed value and Kalman gain is skipped, the radial diffusion velocity at the current time is forcibly set to 0, the radial distance is kept as the posterior state estimate of the previous time, and the posterior estimation error covariance matrix is ​​set to be equal to the posterior estimation error covariance matrix of the previous time plus the process noise covariance matrix; finally, the diffusion distance estimate of each grouting hole at each sampling time is output.

4. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 3, characterized in that, In step S3, the process of spatial geometric subdivision and interferometric quantization specifically includes: At the beginning of each control cycle, obtain the estimated values ​​of grouting pressure and diffusion distance from the previous sampling time. Based on the grouting pressure of each grouting hole, a weighting factor is calculated. The process of calculating the weighting factor specifically includes: Calculate the maximum and minimum pressure values ​​among all grouting holes, calculate the normalized value for each grouting hole, and set the weighting factor to the normalized value plus 1; Using the planar coordinates of each grouting hole as seed points and weighting factors as weights, the multiplicative weighted Voronoi diagram algorithm is executed to generate the weighted Thiessen polygon region corresponding to each grouting hole; For any point on a two-dimensional plane, divide the actual geometric distance from the point to the grouting hole by the weighting factor of the grouting hole to obtain the weighted distance. Divide each point into the grouting hole that minimizes its weighted distance to obtain the weighted Thiessen polygon region corresponding to each grouting hole. Based on the ordered vertex coordinate sequence of the weighted Thiessen polygon region corresponding to each grouting hole, the area enclosed by the weighted Thiessen polygon region is calculated using the shoelace formula, which serves as the control area of ​​the grouting hole.

5. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 4, characterized in that, The process of calculating the geometric interference area in step S3 specifically includes: In the interference calculation of the buffer warning zone, the weighted Thiessen polygon region of each grouting hole is shifted outward by a preset safe buffer distance to generate a buffer extended polygon. For all adjacent grouting hole pairs, calculate the area of ​​the intersection region of the two buffer extended polygons, where adjacent grouting hole pairs are grouting hole pairs with a common edge in the corresponding weighted Thiessen polygon regions. For each grouting hole, the geometric interference area of ​​the intersection region generated by it and all adjacent grouting holes is summed.

6. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 5, characterized in that, In step S4, the parameter combination optimized by the multi-objective optimization algorithm is grouting pressure and grout dynamic viscosity. The optimization objectives include: maximizing the crack sealing efficiency and maximizing the strength of the age-cured structure; The process of calculating the maximum crack sealing efficiency specifically includes: Predict the grout diffusion distance after one control cycle under the action of candidate grouting pressure and grout dynamic viscosity; The crack sealing efficiency is calculated based on the ratio of the predicted slurry diffusion distance to the target diffusion radius; A penalty factor based on spatial geometric interference is introduced to correct the sealing effectiveness of the fracture section. The correction process specifically includes: The ratio of the geometric interference area to the control area of ​​each grouting hole is used as the spatial penalty factor, and the spatial penalty factor is truncated. The crack section sealing effectiveness is corrected based on the truncated spatial penalty factor to obtain the corrected crack section sealing effectiveness.

7. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 6, characterized in that, The mathematical expression for the strength of the time-cured structure is: ； in, For age-cured structural strength, The first fitting coefficient, The dynamic viscosity of the candidate slurry. For reference viscosity, is the second fitting coefficient.

8. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 7, characterized in that, In step S4, the evolution process of the multi-objective optimization algorithm specifically includes: The population size and number of iterations of the preset evolutionary algorithm are used to select samples from the parent population through a tournament selection method. Simulated binary crossover is performed on the samples with a set probability to generate offspring. Mutation operation is performed on all generated offspring individuals, wherein the mutation operation is determined by the current spatial topology. The process of performing the mutation operation specifically includes: Extract the spatial penalty factor value after the grouting hole is truncated and compare it with the preset penalty threshold; if the truncated spatial penalty factor value is less than the penalty threshold, it is determined that the current grouting hole is in a safe and cooperative state, and regular Gaussian mutation is performed. If the truncated space penalty factor value is greater than or equal to the penalty threshold, the current grouting hole is determined to be in a collision danger state. The directional dynamic compression of the variation space is performed, the upper limit of the allowable variation of the candidate grouting pressure is temporarily lowered, the lower limit of the allowable variation of the candidate grout dynamic viscosity is temporarily raised, and conventional Gaussian variation is performed within the compressed range. After the optimization algorithm completes its iterations, it outputs the Pareto front non-dominated solution set on the optimization objective. All non-dominated solutions on the Pareto front are sorted, and the solution ranked first is extracted as the candidate optimal parameter combination. The evaluation algorithm is called to calculate the stress intensity factor corresponding to the parameter combination. If the stress intensity factor is less than the crack initiation toughness constant of the grouting medium, the parameter combination is taken as the final safe and feasible solution. If the stress intensity factor is greater than or equal to the crack initiation toughness constant of the grouting medium, then the next non-dominated solution is extracted sequentially for verification until a parameter combination that satisfies the stress intensity factor being less than the crack initiation toughness constant of the grouting medium is found. The final grouting pressure and grout dynamic viscosity values ​​are sent to the corresponding grouting unit's programmable logic controller for execution.

9. The collaborative control method for grouting operations in building engineering based on edge computing as described in claim 8, characterized in that, The evaluation algorithm takes the candidate grouting pressure, the continuous grouting time within the control cycle, and the dynamic viscosity of the grout as inputs, calculates the stress intensity factor at the crack tip, and outputs the stress intensity factor. The evaluation algorithm process specifically includes: The cumulative injection volume of the grout under the action of candidate parameters is calculated, and the mathematical expression for the cumulative injection volume is: ； in, For the cumulative injection volume, For the comprehensive flow coefficient, For grouting pressure, For continuous grouting time, The dynamic viscosity of the slurry; The fracture half-length is calculated based on the cumulative injected volume, and the stress intensity factor is calculated based on the grouting pressure and the fracture half-length. The mathematical expression for the stress intensity factor is as follows: ； in, Stress intensity factor For grouting pressure, The length of the fissure is half its length.

10. A collaborative control system for grouting operations in building engineering based on edge computing, applied in the collaborative control method for grouting operations in building engineering based on edge computing as described in any one of claims 1-9, characterized in that, It includes a data acquisition module, an estimation module, a data segmentation module, an optimization module, and a verification module; The acquisition module is used to collect construction data, grouting parameters, and engineering parameters; The estimation module is used to perform a state estimation algorithm on each grouting hole based on the construction data, grouting parameters and engineering parameters, and obtain the estimated value of the diffusion distance of the grout diffusion front at the current moment; The segmentation module is used to perform spatial geometric segmentation and interference quantization based on the construction data, and to calculate the control area and geometric interference area of ​​each grouting hole; The optimization module is used to solve for the optimal grouting parameter combination for the next control cycle based on the estimated control area, geometric interference area, and diffusion distance using a multi-objective optimization algorithm. The verification module is used to perform safety verification on the optimal grouting parameter combination and send out the parameters that pass the verification for execution.

Images