Grouting reinforcement correction method for high-rise building structure

Abstract

The invention provides a grouting reinforcement deviation rectifying method for a high-rise building structure, and relates to the technical field of building engineering. The grouting reinforcement correction method for the high-rise building structure comprises the steps of S1, based on the current situation of the high-rise building structure, S2, based on an initial data set, S3, based on a preprocessed data model, S4, based on an optimized grouting scheme, S5, based on data collected by a real-time monitoring system, S6, based on dynamic adjustment parameters, and S7, based on grouting execution records and the real-time monitoring system data; s8, based on real-time monitoring feedback; and S9, based on an effect evaluation report. Building inclination, foundation soil texture and environment data are automatically collected by adopting unmanned aerial vehicle laser scanning and a sensor network, and intelligent analysis and mode recognition are performed in combination with data cleaning and a convolutional neural network, so that high-precision modeling of a building state is realized, and data reliability and evaluation accuracy are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of building engineering technology, specifically to a grouting reinforcement and correction method for high-rise building structures. Background Technology

[0002] Architectural engineering technology is an important branch of civil engineering, focusing on the design, construction, maintenance, repair, and renovation of buildings. Its core objective is to ensure the safety, durability, and functionality of structures. This field involves the interdisciplinary integration of structural mechanics, materials science, geological engineering, and construction management. It primarily addresses the stability issues of high-rise buildings, historical buildings, or existing structures. Through assessment, reinforcement, correction, and monitoring (such as grouting, carbon fiber reinforcement, and jacking technology), it addresses damage caused by foundation settlement, material aging, load changes, or natural disasters to improve seismic performance, extend service life, and meet modern safety standards. Standardization is a key technological direction for ensuring the sustainable operation of urban infrastructure. One method for grouting reinforcement and correction of high-rise building structures involves injecting specific grouts (such as cement-based or chemical grouts) into the building foundation or structure to fill voids, strengthen the soil or concrete, and adjust the building's horizontal position using a controllable grouting process. Its purpose is to solve the tilting and structural instability problems of high-rise buildings caused by uneven foundation settlement, external loads, or environmental factors. By enhancing the overall bearing capacity, correcting the tilt, and reducing crack deformation, it ultimately restores the building to safe use conditions and prevents further damage.

[0003] Existing technologies rely on manual surveys and simple instruments for data collection, resulting in limited data coverage and susceptibility to subjective errors, affecting the comprehensiveness and accuracy of the assessment. Grouting schemes are mostly based on experience-based decisions, lacking systematic simulation and optimization processes, often leading to unreasonable grouting point layouts or inaccurate grout dosages, resulting in unstable corrective effects and significant resource waste. Insufficient real-time monitoring means it is difficult to monitor and adjust dynamic changes during construction in real time, resulting in slow corrective response and potentially delaying the optimal intervention time. Traditional grouting materials have limited performance and cannot adapt to environmental changes, leading to poor durability of reinforcement effects. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a grouting reinforcement and correction method for high-rise building structures. This method solves the problems of existing technologies relying on manual surveys and simple instruments for data collection, which results in limited data coverage and is prone to subjective errors, affecting the comprehensiveness and accuracy of the assessment. Furthermore, grouting schemes are often based on empirical decisions and lack systematic simulation and optimization processes, often leading to unreasonable grouting point layouts or inaccurate grout dosages, resulting in unstable correction effects and serious resource waste.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a grouting reinforcement and correction method for high-rise building structures, comprising the following steps:

[0006] S1: Based on the current structural status of high-rise buildings, UAV laser scanning and sensor networks are used to collect building tilt data, foundation soil data and environmental data, and the data are initially integrated through data acquisition algorithms to generate an initial dataset;

[0007] S2: Based on the initial dataset, data cleaning and feature extraction algorithms are used to preprocess the data, remove noise and outliers, and data analysis and pattern recognition are performed through convolutional neural networks to construct a building tilt model and generate a preprocessed data model.

[0008] S3: Based on the preprocessed data model, the building information model and finite element analysis integrated platform are used to simulate the stress distribution and displacement changes after grouting. The layout of grouting points, grout type and injection sequence are optimized by genetic algorithm to minimize tilting error and generate an optimized grouting scheme.

[0009] S4: Based on the optimized grouting scheme, deploy an IoT sensor network to achieve real-time data transmission, and set the monitoring frequency and alarm threshold through network configuration algorithms to ensure that the system covers key areas and generate a real-time monitoring system;

[0010] S5: Based on data collected by the real-time monitoring system, the model predictive control algorithm is used to dynamically predict the building tilt trend, and the grouting parameters are adjusted through proportional-integral-derivative control to respond to real-time changes and generate dynamic adjustment parameters.

[0011] S6: Based on dynamically adjusted parameters, the grouting equipment is controlled to perform grouting operations. Intelligent responsive grout is used and precise injection is carried out through an adaptive injection system. Priority is given to implementation in settlement areas, and grouting execution records are generated.

[0012] S7: Based on grouting execution records and real-time monitoring system data, a closed-loop control algorithm is adopted to monitor grouting effect and tilt change in real time, and optimize the injection process through feedback loop, calibrate deviations in a timely manner, and generate real-time monitoring feedback.

[0013] S8: Based on real-time monitoring feedback, a multi-objective optimization algorithm is used to evaluate the correction effect, and a comprehensive evaluation report is generated through Pareto frontier analysis.

[0014] S9: Based on the effect evaluation report, use artificial intelligence algorithms to generate a final corrective summary and subsequent maintenance suggestions, and output visualized results through a report generation tool to ensure the continuous optimization of the solution and generate the final optimized solution.

[0015] Preferably, generating the initial dataset based on S1 includes the following steps:

[0016] S101: Based on the need for high-rise building correction, a drone equipped with a lidar and a high-definition camera module is used to perform three-dimensional laser scanning of the building facade and surrounding environment, and a high-precision building geometric model is obtained through a point cloud data generation algorithm to generate the initial point cloud model of the building.

[0017] S102: Based on the initial point cloud model of the building, a key point sensor network is deployed inside the building and on the foundation. The real-time deformation and tilt data of the structure are automatically collected through the wireless sensor network protocol to generate a real-time structural monitoring dataset.

[0018] S103: Based on the real-time structural monitoring dataset, the foundation exploration is carried out using ground-penetrating radar and standard penetration test methods to obtain soil layer distribution, density and groundwater level data, and then the data are digitized through geological statistical analysis to generate a foundation geological attribute database.

[0019] S104: Based on the initial point cloud model of the building, the real-time monitoring dataset of the structure, and the geological attribute database of the ground, a data fusion algorithm is used to perform spatiotemporal registration and integration of multi-source heterogeneous data, eliminate noise and conflicts, form a basic data package in a unified format, and generate the initial dataset.

[0020] Preferably, generating a preprocessed data model based on S2 includes the following steps:

[0021] S201: Based on the initial dataset, a statistical outlier detection algorithm is used to identify and remove obvious errors and noisy data points generated during the collection process, generating a cleaned dataset;

[0022] S202: Based on the cleaned dataset, the principal component analysis algorithm is used to reduce the dimensionality of the high-dimensional monitoring data, extract the main feature components that best represent the building tilt and settlement trend, and generate the building settlement main feature vector.

[0023] S203: Based on the main feature vector of building settlement, a convolutional neural network algorithm is used to perform deep learning on historical and real-time tilt patterns to identify settlement patterns, rates and potential risk areas, and generate a building tilt pattern identification report.

[0024] S204: The preprocessed data model is generated by combining the cleaned dataset, the main feature vector of building settlement, and the building tilt pattern recognition report, and using data standardization methods to normalize all features to the same dimension and encapsulate them into standard data objects that can be called by the model.

[0025] Preferably, the S3-based optimized grouting scheme includes the following steps:

[0026] S301: Based on the preprocessed data model, a high-precision three-dimensional digital twin model containing the building structure, foundation soil layer and sensor points is constructed using building information modeling software to generate a building-foundation digital twin model.

[0027] S302: Based on the building-foundation digital twin model, the finite element analysis algorithm is used to perform mechanical simulation under different grouting parameters, calculate stress redistribution and expected displacement field, and generate a set of grouting condition simulation results;

[0028] S303: Based on the grouting condition simulation result set, a multi-objective genetic algorithm is adopted, with "maximizing the correction effect", "minimizing the material consumption" and "minimizing the construction time" as the objective functions, to perform multiple rounds of iterative optimization, search for the optimal solution set, and generate the Pareto optimal solution set;

[0029] S304: Based on the Pareto optimal solution set, a decision support system is used to combine engineering experience and cost constraints to select a final implementation plan from the solution set, clarify the grouting hole location, sequence, grout ratio and pressure parameters, and generate an optimized grouting plan.

[0030] Preferably, the S4-based real-time monitoring system includes the following steps:

[0031] S401: Based on the key monitoring areas identified in the optimized grouting scheme, an IoT device deployment strategy is adopted to supplement the deployment of wireless inclinometers, micromanometers, and earth pressure cell sensors to form a high-density monitoring network and generate an enhanced sensor network deployment map.

[0032] S402: Based on the enhanced sensor network deployment diagram, a wireless networking protocol is used to complete the self-organizing network connection of all sensors, and a time division multiple access scheduling algorithm is used to configure the data acquisition and transmission frequency and generate a real-time data acquisition scheduling table.

[0033] S403: Based on the real-time data acquisition scheduling table, cloud-based data receiving and storage services are used to build a data pipeline to receive, verify, and persistently store massive amounts of incoming real-time monitoring data, generating a cloud-based raw monitoring database.

[0034] S404: Based on the cloud-based raw monitoring database, it uses a rule engine to set various alarm thresholds and configures data visualization dashboards and early warning information push mechanisms to generate a real-time monitoring system.

[0035] Preferably, the generation of dynamically adjusted parameters based on S5 includes the following steps:

[0036] S501: Based on the real-time data stream pushed by the real-time monitoring system, the Kalman filter algorithm is used for data fusion and filtering to estimate the optimal value of the building tilt state, reduce the impact of measurement noise, and generate the optimal value of the building state.

[0037] S502: Based on the optimal estimated value of the building status and the expected target in the optimized grouting scheme, the model predictive control algorithm is used to predict the building tilt trend in the future period and generate the tilt trend prediction value.

[0038] S503: Based on the tilt trend prediction value, the proportional-integral-derivative control algorithm is used to calculate the adjustment amount required for the current grouting parameters so that the prediction value approaches the target value, and a set of grouting parameter adjustment instructions is generated.

[0039] S504: Sends the grouting parameter adjustment instruction set to the grouting equipment controller, drives the intelligent grouting pump and valve actuator to make precise adjustments, realizes closed-loop adaptive control of the grouting process, and generates dynamic adjustment parameters.

[0040] Preferably, the generation of grouting execution records based on S6 includes the following steps:

[0041] S601: Based on dynamically adjusted parameters, a programmable logic controller is used to control the start-up, shutdown, pressure, and speed of the high-pressure grouting pump, ensuring that the output parameters are consistent with the instructions and generating pumping control signals;

[0042] S602: Based on the pumping control signal, an intelligent grout delivery system is adopted to automatically mix cement, chemical grout and nano-additives according to the proportion in the optimized grouting scheme to prepare intelligent responsive grout and generate customized grout.

[0043] S603: The customized grout is delivered to the grouting holes specified in the optimized grouting scheme through a distributed grouting pipeline system, and the injection status of each hole is monitored in real time using flow meters and pressure sensors to generate real-time data on hole injection.

[0044] S604: Based on pumping control signals and real-time data of hole injection, a data log recording algorithm is adopted, using timestamps as indexes, to fully record the grouting parameters, equipment status and injection volume at each moment, and generate grouting execution records.

[0045] Preferably, generating real-time monitoring feedback based on S7 includes the following steps:

[0046] S701: Based on the continuous data stream of grouting execution records and real-time monitoring system, a time series comparison algorithm is used to compare the actual settlement / lifting data after grouting with the tilt trend prediction value in real time, and generate a real-time comparison report of grouting effect.

[0047] S702: Based on the real-time comparison report of grouting effect, the deviation analysis algorithm is used to calculate the error between the actual effect and the expected effect, locate the grouting area with poor effect or abnormality, and generate a grouting deviation analysis diagram.

[0048] S703: Feeds back the grouting deviation analysis diagram to the model predictive control algorithm of S5 as the input for a new round of predictive control, corrects the predictive model parameters in real time, optimizes the accuracy of subsequent control commands, and generates model parameter correction amounts.

[0049] S704: Based on the model parameter correction amount, a control loop iterative mechanism is adopted to dynamically update the control strategy and generate decision suggestions on whether to supplement grouting or adjust the focus, and generate real-time monitoring feedback.

[0050] Preferably, the S8-based generation effect evaluation report includes the following steps:

[0051] S801: Based on real-time monitoring feedback and grouting execution records, a data aggregation algorithm is used to extract key performance indicators such as the final inclination, total grout usage, and total man-hours after the entire construction cycle, and generate a set of performance indicators for the correction project.

[0052] S802: Based on the performance index set of the correction project, a multi-attribute decision analysis method is adopted, and a comprehensive evaluation system is constructed by combining the dimensions of economic efficiency, environmental impact and social benefits, and a multi-dimensional evaluation matrix is ​​generated.

[0053] S803: Based on a multi-dimensional evaluation matrix, a particle swarm optimization algorithm is used for non-dominated sorting to find a set of solutions that perform well under multiple objectives, forming a Pareto front and generating a comprehensive evaluation Pareto front.

[0054] S804: Based on a comprehensive assessment of the Pareto frontier, a report generation engine is used to automatically generate a structured assessment report, which includes key findings, scores for each indicator, cost-benefit analysis, and improvement suggestions, thus generating an effectiveness assessment report.

[0055] Preferably, the generation of the final optimized solution based on S9 includes the following steps:

[0056] S901: Based on the effectiveness evaluation report, use natural language processing technology to extract key conclusions, successful experiences and existing problems from the report, and generate a key information extraction summary;

[0057] S902: Based on key information extraction, a summary is extracted, and reinforcement learning is used to simulate the long-term effects of different subsequent maintenance strategies. The agent is trained to learn the optimal long-term maintenance scheme and generate simulation results of long-term maintenance strategies.

[0058] S903: Based on the simulation results of long-term maintenance strategy, the case reasoning method is used to store all data, models and results of this project into the knowledge base to provide reference for similar projects in the future and generate grouting correction knowledge base update records;

[0059] S904: Integrates key information extraction summaries, long-term maintenance strategy simulation results, and grouting correction knowledge base update records, and uses intelligent document generation technology to output the final summary report, operation and maintenance manual, and sustainable optimization suggestions, generating the final optimization plan.

[0060] This invention provides a grouting reinforcement and correction method for high-rise building structures. It has the following beneficial effects:

[0061] This invention utilizes drone laser scanning and sensor networks to automatically collect data on building tilt, foundation soil quality, and the environment. Combined with data cleaning and convolutional neural networks for intelligent analysis and pattern recognition, it achieves high-precision modeling of the building's condition, significantly improving data reliability and assessment accuracy. Digital simulation is performed using Building Information Modeling (BIM) and finite element analysis, and grouting parameters are optimized through genetic algorithms, enhancing the scientific rigor and predictive capabilities of the scheme and reducing trial-and-error costs. Real-time monitoring is achieved through the deployment of an IoT sensor network. Model predictive control and proportional-integral-derivative (PID) control are combined to dynamically adjust the grouting process, ensuring rapid response and precise correction to changing environments, improving construction efficiency and adaptability. Intelligent responsive grout and an adaptive injection system are used to perform precise operations, enhancing grouting effects and durability. Continuous optimization of operations is achieved through closed-loop control algorithms and multi-objective optimization assessments, and sustainable maintenance recommendations are generated based on artificial intelligence. Overall, this improves correction accuracy, resource utilization, and project sustainability while reducing the risk of human intervention. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the main steps of the present invention;

[0063] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0064] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0065] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0066] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0067] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0068] Figure 7 This is a detailed schematic diagram of S6 of the present invention;

[0069] Figure 8 This is a detailed schematic diagram of S7 of the present invention;

[0070] Figure 9 This is a detailed schematic diagram of S8 of the present invention;

[0071] Figure 10 This is a detailed schematic diagram of S9 of the present invention. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] Example:

[0074] like Figures 1-10 As shown, this embodiment of the invention provides a grouting reinforcement and correction method for high-rise building structures, including the following steps:

[0075] S1: Based on the current structural status of high-rise buildings, UAV laser scanning and sensor networks are used to collect building tilt data, foundation soil data and environmental data, and the data are initially integrated through data acquisition algorithms to generate an initial dataset;

[0076] S2: Based on the initial dataset, data cleaning and feature extraction algorithms are used to preprocess the data, remove noise and outliers, and data analysis and pattern recognition are performed through convolutional neural networks to construct a building tilt model and generate a preprocessed data model.

[0077] S3: Based on the preprocessed data model, the building information model and finite element analysis integrated platform are used to simulate the stress distribution and displacement changes after grouting. The layout of grouting points, grout type and injection sequence are optimized by genetic algorithm to minimize tilting error and generate an optimized grouting scheme.

[0078] S4: Based on the optimized grouting scheme, deploy an IoT sensor network to achieve real-time data transmission, and set the monitoring frequency and alarm threshold through network configuration algorithms to ensure that the system covers key areas and generate a real-time monitoring system;

[0079] S5: Based on data collected by the real-time monitoring system, the model predictive control algorithm is used to dynamically predict the building tilt trend, and the grouting parameters are adjusted through proportional-integral-derivative control to respond to real-time changes and generate dynamic adjustment parameters.

[0080] S6: Based on dynamically adjusted parameters, the grouting equipment is controlled to perform grouting operations. Intelligent responsive grout is used and precise injection is carried out through an adaptive injection system. Priority is given to implementation in settlement areas, and grouting execution records are generated.

[0081] S7: Based on grouting execution records and real-time monitoring system data, a closed-loop control algorithm is adopted to monitor grouting effect and tilt change in real time, and optimize the injection process through feedback loop, calibrate deviations in a timely manner, and generate real-time monitoring feedback.

[0082] S8: Based on real-time monitoring feedback, a multi-objective optimization algorithm is used to evaluate the correction effect, and a comprehensive evaluation report is generated through Pareto frontier analysis.

[0083] S9: Based on the effect evaluation report, use artificial intelligence algorithms to generate a final corrective summary and subsequent maintenance suggestions, and output visualized results through a report generation tool to ensure the continuous optimization of the solution and generate the final optimized solution.

[0084] Generating the initial dataset based on S1 involves the following steps:

[0085] S101: Based on the need for high-rise building correction, a drone equipped with a lidar and a high-definition camera module is used to perform three-dimensional laser scanning of the building facade and surrounding environment, and a high-precision building geometric model is obtained through a point cloud data generation algorithm to generate the initial point cloud model of the building.

[0086] To address the need for high-rise building alignment correction, a DJI Matrice 300 RTK drone equipped with a Zenmuse L1 LiDAR module was used. The drone was set to fly at an altitude of 50 meters and a speed of 5 meters per second to perform a 3D laser scan of the building facade. The scan path was pre-programmed as a spiral descent around the building. The LiDAR emitted pulsed lasers, measured distances, and recorded return times. Combined with GPS and IMU data, point cloud data was acquired. For example, for a building height of 100 meters and a scanned area of ​​2000 square meters, approximately 5 million point cloud data points were obtained. The point cloud density was calculated by dividing the total number of points by the scanned area: 5,000,000 / 2000 = 2500 points / m². CloudCompare software was used for point cloud processing, applying a statistical outlier removal algorithm to calculate the average distance from each point in the point cloud to its neighbors, obtaining the mean μ and standard deviation σ. A threshold of μ ± σ was set. 2σ, for example, if μ is 10 meters and σ is 0.5 meters, then points whose distance values ​​are not in the range of 9.0 to 11.0 meters are removed. After processing, the point cloud is registered and triangulated to generate a high-precision three-dimensional mesh model with vertex position accuracy error of less than 1 cm, and the initial point cloud model of the building is generated.

[0087] S102: Based on the initial point cloud model of the building, a key point sensor network is deployed inside the building and on the foundation. The real-time deformation and tilt data of the structure are automatically collected through the wireless sensor network protocol to generate a real-time structural monitoring dataset.

[0088] Based on the initial point cloud model of the building, wireless tilt sensors and strain gauges are deployed at key locations such as corners and the foundation. The sensor type is SICK DT1000 tiltmeter, with a sampling frequency of 10Hz. They are networked via the ZigBee wireless protocol to automatically collect real-time deformation data. The tilt data is recorded as angle values, for example, the sensor reading shows a tilt angle of 0.5 degrees. The data is transmitted to the central server through the gateway. The data packet includes a timestamp, sensor ID, and measurement value. Data integrity is checked using data verification algorithms, such as CRC check, to ensure no data loss or errors. The acquisition cycle is once per minute for 24 hours, generating a real-time structural monitoring dataset containing time-series data.

[0089] S103: Based on the real-time structural monitoring dataset, the foundation exploration is carried out using ground-penetrating radar and standard penetration test methods to obtain soil layer distribution, density and groundwater level data, and then the data are digitized through geological statistical analysis to generate a foundation geological attribute database.

[0090] Based on the real-time structural monitoring dataset, ground exploration was conducted using ground-penetrating radar equipment such as the GSSI SIR-4000. The radar antenna frequency was 100MHz, and the scanning depth was 20 meters to acquire soil reflection signals. After filtering and gain adjustment, soil interfaces were identified. For example, a clay layer with high reflection intensity was found at a depth of 5 meters. At the same time, standard penetration tests were conducted using a 63.5kg hammer to record the number of blows (N) per 30cm of penetration. For example, at point A, the N value was 15, indicating medium density. Groundwater level was measured through boreholes and water level gauges at a depth of 3 meters. The data was digitized and stored in tables. The soil layer distribution was represented by a layered model, and the thickness and properties of each layer were recorded to generate a ground geological property database.

[0091] S104: Based on the initial point cloud model of the building, the real-time monitoring dataset of the structure, and the geological attribute database of the ground, a data fusion algorithm is used to perform spatiotemporal registration and integration of multi-source heterogeneous data, eliminate noise and conflicts, form a basic data package in a unified format, and generate the initial dataset.

[0092] Based on the initial point cloud model of the building, the real-time structural monitoring dataset, and the geological attribute database, a data fusion algorithm such as Kalman filtering is used to perform spatiotemporal registration of the multi-source data. First, the point cloud model is aligned with the monitoring data in time using timestamp matching. Then, spatial registration is performed through coordinate transformation, such as moving the sensor coordinate system to the point cloud coordinate system. Data conflict resolution uses a weighted average, with weights based on data accuracy: point cloud data weight 0.6, sensor data weight 0.3, and geological data weight 0.1. Noise removal is performed using a low-pass filter with a cutoff frequency of 0.1Hz. A unified format CSV file containing all integrated data is generated, thus creating the initial dataset.

[0093] The steps involved in generating the preprocessed data model based on S2 are as follows:

[0094] S201: Based on the initial dataset, a statistical outlier detection algorithm is used to identify and remove obvious errors and noisy data points generated during the collection process, generating a cleaned dataset;

[0095] Based on the initial dataset, a statistical outlier detection algorithm is used to calculate the outlier detection value for each data point. , Where X is the data value, μ is the mean, and σ is the standard deviation. A threshold is set: |Z-score| > 3 is considered an anomaly. For example, if the tilt data has μ of 0.4 degrees and σ of 0.1 degrees, it will be removed. Points with a value greater than 3 or less than -3, such as points with a value of 0.8 degrees. The data was removed, and after cleaning, a moving average smoothing was applied with a window size of 5 points to generate the cleaned dataset.

[0096] S202: Based on the cleaned dataset, the principal component analysis algorithm is used to reduce the dimensionality of the high-dimensional monitoring data, extract the main feature components that best represent the building tilt and settlement trend, and generate the building settlement main feature vector.

[0097] Based on the cleaned dataset, principal component analysis algorithm is used to calculate the covariance matrix of the data and extract principal components through eigenvalue decomposition. For example, if the dataset has 10 variables, the variance contribution rate of the first principal component is 80% after calculation. The first 3 principal components are retained, and the eigenvectors represent the tilt trend. After dimensionality reduction, the data is projected onto the principal component space to generate the main eigenvectors of building settlement.

[0098] S203: Based on the main feature vector of building settlement, a convolutional neural network algorithm is used to perform deep learning on historical and real-time tilt patterns to identify settlement patterns, rates and potential risk areas, and generate a building tilt pattern identification report.

[0099] Based on the main feature vector of building settlement, a convolutional neural network algorithm is adopted. The network structure includes convolutional layers, pooling layers, and fully connected layers. The input is time series data with a sequence length of 100 time points, a convolutional kernel size of 3, and a stride of 1. Historical data is used to predict tilt patterns. For example, if the input is data from the past week, the output will show the future trend. The settlement rate is identified, such as 0.1 mm per day. Risk areas are identified by outputting a probability map. A probability > 0.7 indicates high risk. A building tilt pattern recognition report is generated.

[0100] S204: The preprocessed data model is generated by combining the cleaned dataset, the main feature vector of building settlement, and the building tilt pattern recognition report, and using data standardization methods to normalize all features to the same dimension and encapsulate them into standard data objects that can be called by the model.

[0101] Based on the comprehensive cleaned dataset, the main feature vector of building settlement, and the building tilt pattern recognition report, a data standardization method was used to scale each feature to the range [0,1]. The formula is as follows: ,in and To determine the minimum and maximum values ​​of features, for example, if the tilt angle X_min = 0 degrees and X_max = 1 degree, then the value 0.5 degrees is standardized to 0.5. All features are encapsulated into JSON objects to generate a preprocessed data model.

[0102] The optimized grouting scheme for S3 includes the following steps:

[0103] S301: Based on the preprocessed data model, a high-precision three-dimensional digital twin model containing the building structure, foundation soil layer and sensor points is constructed using building information modeling software to generate a building-foundation digital twin model.

[0104] Based on the preprocessed data model, point cloud data and geological data are imported using building information modeling software such as Autodesk Revit to construct a three-dimensional digital model. The model includes building structural layers, foundation soil layers, and sensor points. The soil layer thickness is set according to the geological data, for example, the clay layer thickness is 2 meters. The model accuracy is LOD 400, generating a building-foundation digital twin model.

[0105] S302: Based on the building-foundation digital twin model, the finite element analysis algorithm is used to perform mechanical simulation under different grouting parameters, calculate stress redistribution and expected displacement field, and generate a set of grouting condition simulation results;

[0106] Based on a building-foundation digital twin model, finite element analysis algorithms such as ANSYS software were used. Grouting parameters were set, with grouting pressure ranging from 0.1 MPa to 0.5 MPa in increments of 0.1 MPa. Hooke's law was used to simulate stress distribution. Where E is the elastic modulus and ε is the strain, for example, E=30GPa for concrete. Calculate the displacement field, output the maximum displacement value, and generate a set of simulation results for grouting conditions.

[0107] S303: Based on the grouting condition simulation result set, a multi-objective genetic algorithm is adopted, with "maximizing the correction effect", "minimizing the material consumption" and "minimizing the construction time" as the objective functions, to perform multiple rounds of iterative optimization, search for the optimal solution set, and generate the Pareto optimal solution set;

[0108] Based on the simulation results set of grouting conditions, a multi-objective genetic algorithm is adopted. The objective function is f1, which is the correction effect and the amount of tilt reduction; f2, which is the material usage and grout volume; and f3, which is the construction time. The population size is 100, the iteration is 50 generations, the crossover probability is 0.8, and the mutation probability is 0.1. The Pareto front is calculated. For example, for solution A: f1=90%, f2=10m³, f3=50h, and for solution B: f1=85%, f2=8m³, f3=40h, a Pareto optimal solution set is generated.

[0109] S304: Based on the Pareto optimal solution set, a decision support system is used to combine engineering experience and cost constraints to select a final implementation plan from the solution set, clarify the grouting hole location, sequence, grout ratio and pressure parameters, and generate an optimized grouting plan.

[0110] Based on the Pareto optimal solution set, a decision support system is used, with input weights for engineering experience (0.5), economic efficiency (0.3), environmental impact (0.2), and safety (0.2). The comprehensive score for each solution is calculated. , Normalize f1 to [0,1], select the solution with the highest score, clarify the grouting hole coordinates, sequence order, grout mix ratio cement:water = 1:0.4, pressure parameter 0.3MPa, and generate an optimized grouting scheme.

[0111] The generation of a real-time monitoring system based on S4 includes the following steps:

[0112] S401: Based on the key monitoring areas identified in the optimized grouting scheme, an IoT device deployment strategy is adopted to supplement the deployment of wireless inclinometers, micromanometers, and earth pressure cell sensors to form a high-density monitoring network and generate an enhanced sensor network deployment map.

[0113] Based on the key monitoring areas identified in the optimized grouting scheme, an IoT device deployment strategy was adopted to supplement the settlement area with wireless inclinometers, micromanometers, and earth pressure cell sensors. The sensors were spaced 10 meters apart, with a total of 50 sensors connected via the LoRaWAN protocol in a star topology, generating an enhanced sensor network deployment diagram.

[0114] S402: Based on the enhanced sensor network deployment diagram, a wireless networking protocol is used to complete the self-organizing network connection of all sensors, and a time division multiple access scheduling algorithm is used to configure the data acquisition and transmission frequency and generate a real-time data acquisition scheduling table.

[0115] Based on the enhanced sensor network deployment diagram, a wireless networking protocol is adopted, and network parameters such as frequency 868MHz and bandwidth 125kHz are configured. The time division multiple access scheduling algorithm is used, and each sensor is allocated a time slot to transmit data once per second, with a slot length of 100ms, to generate a real-time data acquisition scheduling table.

[0116] S403: Based on the real-time data acquisition scheduling table, cloud-based data receiving and storage services are used to build a data pipeline to receive, verify, and persistently store massive amounts of incoming real-time monitoring data, generating a cloud-based raw monitoring database.

[0117] Based on the real-time data acquisition schedule table, cloud-based data receiving services such as AWS IoT Core are used to set up a data pipeline. Data verification uses the checksum algorithm and is stored in a MySQL database. The table structure includes timestamp, sensor_id, and value, generating a raw monitoring database in the cloud.

[0118] S404: Based on the cloud-based raw monitoring database, it uses a rule engine to set various alarm thresholds and configures data visualization dashboards and early warning information push mechanisms to generate a real-time monitoring system.

[0119] Based on the cloud-based raw monitoring database, an alarm threshold is set using a rule engine: tilt threshold of 0.3 degrees and pressure threshold of 0.5 MPa. The rule is IF tilt > 0.3 THEN alert. A visual dashboard is configured using Grafana, and alerts are pushed via SMS API, generating a real-time monitoring system.

[0120] The generation of dynamically adjusted parameters based on S5 includes the following steps:

[0121] S501: Based on the real-time data stream pushed by the real-time monitoring system, the Kalman filter algorithm is used for data fusion and filtering to estimate the optimal value of the building tilt state, reduce the impact of measurement noise, and generate the optimal value of the building state.

[0122] Based on the real-time data stream pushed by the real-time monitoring system, a Kalman filter algorithm is used, with the tilt angle as the state variable, measurement noise covariance R=0.01, process noise covariance Q=0.001, and prediction steps. Update steps Where A=1, H=1, u=0, estimate the tilt state, for example, the current estimate is 0.25 degrees, and generate the optimal estimate of the building state.

[0123] S502: Based on the optimal estimated value of the building status and the expected target in the optimized grouting scheme, the model predictive control algorithm is used to predict the building tilt trend in the future period and generate the tilt trend prediction value.

[0124] Based on the optimal estimated value of the building state and the expected target in the optimized grouting scheme, a model predictive control algorithm is adopted, with 10 steps in the prediction time domain and 5 steps in the control time domain. The objective function minimizes the tilt error, and the QP solver is used to solve the optimization problem to generate the tilt trend prediction value.

[0125] S503: Based on the tilt trend prediction value, the proportional-integral-derivative control algorithm is used to calculate the adjustment amount required for the current grouting parameters so that the prediction value approaches the target value, and a set of grouting parameter adjustment instructions is generated.

[0126] Based on the predicted tilt trend, a proportional-integral-derivative (PID) control algorithm is adopted, with Kp=0.5, Ki=0.1, Kd=0.2, and the error e = target - prediction. The control output is... Calculate the grouting pressure adjustment amount, for example, adjust by +0.05MPa, and generate a set of grouting parameter adjustment instructions.

[0127] S504: Sends the grouting parameter adjustment instruction set to the grouting equipment controller, drives the intelligent grouting pump and valve actuator to make precise adjustments, realizes closed-loop adaptive control of the grouting process, and generates dynamic adjustment parameters.

[0128] The grouting parameter adjustment instruction set is sent to the grouting equipment controller, which is a PLC such as Siemens S7-1200. The controller outputs an analog signal of 4-20mA to drive the pump and valve, adjust the pressure to the set value, realize closed-loop control, and generate dynamic adjustment parameters.

[0129] The generation of grouting execution records based on S6 includes the following steps:

[0130] S601: Based on dynamically adjusted parameters, a programmable logic controller is used to control the start-up, shutdown, pressure, and speed of the high-pressure grouting pump, ensuring that the output parameters are consistent with the instructions and generating pumping control signals;

[0131] Based on dynamically adjusted parameters, a programmable logic controller (PLC) is used. The programming logic is IF pressure_setpoint> current_pressure THEN increase_pump_speed. Pump speed control is achieved through PID control and pressure feedback sensor readings to generate pumping control signals.

[0132] S602: Based on the pumping control signal, an intelligent grout delivery system is adopted to automatically mix cement, chemical grout and nano-additives according to the proportion in the optimized grouting scheme to prepare intelligent responsive grout and generate customized grout.

[0133] Based on the pumping control signal, an intelligent slurry delivery system is adopted. The proportion of ingredients is optimized according to the following scheme: 100kg of cement, 40kg of water, and 5kg of nano-additives. The mixing time is 5 minutes, and the slurry density test is 1.8g / cm³, thus generating a customized slurry.

[0134] S603: The customized grout is delivered to the grouting holes specified in the optimized grouting scheme through a distributed grouting pipeline system, and the injection status of each hole is monitored in real time using flow meters and pressure sensors to generate real-time data on hole injection.

[0135] Customized grout is pumped through a distributed grouting pipeline system at a flow rate of 10L / min and a pressure of 0.3MPa to the designated borehole location. The borehole location coordinates are read from the plan and monitored using flow meters and pressure sensors. Data is recorded once per second to generate real-time data on borehole injection.

[0136] S604: Based on pumping control signals and real-time data of hole injection, a data log recording algorithm is adopted, using timestamps as indexes, to fully record the grouting parameters, equipment status and injection volume at each moment, and generate grouting execution records.

[0137] Based on pumping control signals and real-time data of hole injection, a data logging algorithm is adopted. The log format is CSV, with columns including time, pressure, flow rate, and volume, and timestamp accuracy in milliseconds, to generate grouting execution records.

[0138] Generating real-time monitoring feedback based on S7 includes the following steps:

[0139] S701: Based on the continuous data stream of grouting execution records and real-time monitoring system, a time series comparison algorithm is used to compare the actual settlement / lifting data after grouting with the tilt trend prediction value in real time, and generate a real-time comparison report of grouting effect.

[0140] Based on the continuous data stream from the grouting execution records and the real-time monitoring system, a time series comparison algorithm is used to calculate the difference between the actual settlement value and the predicted value. The absolute values ​​of the differences are averaged, for example, if the average difference is 0.02 mm, to generate a real-time comparison report of the grouting effect.

[0141] S702: Based on the real-time comparison report of grouting effect, the deviation analysis algorithm is used to calculate the error between the actual effect and the expected effect, locate the grouting area with poor effect or abnormality, and generate a grouting deviation analysis diagram.

[0142] Based on the real-time comparison report of grouting effects, a deviation analysis algorithm is used to analyze the deviation. Set a deviation threshold of 10%. Areas exceeding the threshold are marked as abnormal. For example, if the deviation in area A is 15%, it will be located and a grouting deviation analysis chart will be generated.

[0143] S703: Feeds back the grouting deviation analysis diagram to the model predictive control algorithm of S5 as the input for a new round of predictive control, corrects the predictive model parameters in real time, optimizes the accuracy of subsequent control commands, and generates model parameter correction amounts.

[0144] The grouting deviation analysis diagram is fed back to the model predictive control algorithm as input to adjust the model parameters, such as modifying the process noise Q value, rerunning the prediction, and generating the model parameter correction amount.

[0145] S704: Based on the model parameter correction amount, a control loop iterative mechanism is adopted to dynamically update the control strategy and generate decision suggestions on whether to supplement grouting or adjust the focus, and generate real-time monitoring feedback.

[0146] Based on the model parameter correction, a control loop iterative mechanism is adopted to update the control strategy, such as increasing the Kp gain. Decision suggestions are based on the deviation magnitude: IF deviation>10% THEN recommend_additional_injection, generating real-time monitoring feedback.

[0147] The S8-based performance evaluation report includes the following steps:

[0148] S801: Based on real-time monitoring feedback and grouting execution records, a data aggregation algorithm is used to extract key performance indicators such as the final inclination, total grout usage, and total man-hours after the entire construction cycle, and generate a set of performance indicators for the correction project.

[0149] Based on real-time monitoring feedback and grouting execution records, a data aggregation algorithm was used to calculate the total tilt reduction from 0.5 degrees to 0.1 degrees, the total grout usage of 50 m³, and the total working hours of 100 hours, generating a set of performance indicators for the correction project.

[0150] S802: Based on the performance index set of the correction project, a multi-attribute decision analysis method is adopted, and a comprehensive evaluation system is constructed by combining the dimensions of economic efficiency, environmental impact and social benefits, and a multi-dimensional evaluation matrix is ​​generated.

[0151] Based on the performance index set of the corrective engineering project, a multi-attribute decision analysis method is adopted, with attribute weights of 0.4 for economy, 0.3 for environment, and 0.3 for society. The scoring function is linearly weighted to generate a multi-dimensional evaluation matrix.

[0152] S803: Based on a multi-dimensional evaluation matrix, a particle swarm optimization algorithm is used for non-dominated sorting to find a set of solutions that perform well under multiple objectives, forming a Pareto front and generating a comprehensive evaluation Pareto front.

[0153] Based on a multi-dimensional evaluation matrix, a particle swarm optimization algorithm is used with 50 particles and 100 iterations. The objective is to maximize the comprehensive score. The Pareto solution is found by non-dominated sorting, and a comprehensive evaluation Pareto front is generated.

[0154] S804: Based on a comprehensive assessment of the Pareto frontier, a report generation engine is used to automatically generate a structured assessment report, which includes key findings, scores for each indicator, cost-benefit analysis, and improvement suggestions, thus generating an effectiveness assessment report.

[0155] Based on a comprehensive assessment of the Pareto frontier, a report generation engine is used. The template includes chapter results, indicators, and analysis, and automatically populates data to generate an effectiveness evaluation report.

[0156] The final optimized solution based on S9 includes the following steps:

[0157] S901: Based on the effectiveness evaluation report, use natural language processing technology to extract key conclusions, successful experiences and existing problems from the report, and generate a key information extraction summary;

[0158] Based on the effectiveness evaluation report, natural language processing technology is used to extract keywords such as reduced skewness and low cost. The TF-IDF algorithm and word frequency statistics are then used to generate a key information extraction summary.

[0159] S902: Based on key information extraction, a summary is extracted, and reinforcement learning is used to simulate the long-term effects of different subsequent maintenance strategies. The agent is trained to learn the optimal long-term maintenance scheme and generate simulation results of long-term maintenance strategies.

[0160] Based on key information extraction, a reinforcement learning algorithm is used to update the Q-value, simulate the effect of the maintenance strategy, select the optimal strategy after training, and generate simulation results of the long-term maintenance strategy.

[0161] S903: Based on the simulation results of long-term maintenance strategy, the case reasoning method is used to store all data, models and results of this project into the knowledge base to provide reference for similar projects in the future and generate grouting correction knowledge base update records;

[0162] Based on the simulation results of the long-term maintenance strategy, a case-based reasoning method is adopted. The case library stores data on similar projects, retrieves matching cases, updates the knowledge base, and generates grouting correction knowledge base update records.

[0163] S904: Integrates key information extraction summaries, long-term maintenance strategy simulation results, and grouting correction knowledge base update records, and uses intelligent document generation technology to output the final summary report, operation and maintenance manual, and sustainable optimization suggestions, generating the final optimization plan.

[0164] By integrating key information extraction summaries, long-term maintenance strategy simulation results, and grouting correction knowledge base update records, and using intelligent document generation technology, a PDF report is output, including a summary and recommendations, to generate the final optimization plan.

[0165] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A grouting reinforcement and correction method for high-rise building structures, characterized in that, Includes the following steps: S1: Based on the current structural status of high-rise buildings, UAV laser scanning and sensor networks are used to collect building tilt data, foundation soil data and environmental data, and the data are initially integrated through data acquisition algorithms to generate an initial dataset; S2: Based on the initial dataset, data cleaning and feature extraction algorithms are used to preprocess the data, remove noise and outliers, and data analysis and pattern recognition are performed through convolutional neural networks to construct a building tilt model and generate a preprocessed data model. S3: Based on the preprocessed data model, the building information model and finite element analysis integrated platform are used to simulate the stress distribution and displacement changes after grouting. The layout of grouting points, grout type and injection sequence are optimized by genetic algorithm to minimize tilting error and generate an optimized grouting scheme. S4: Based on the optimized grouting scheme, deploy an IoT sensor network to achieve real-time data transmission, and set the monitoring frequency and alarm threshold through network configuration algorithms to ensure that the system covers key areas and generate a real-time monitoring system; S5: Based on data collected by the real-time monitoring system, the model predictive control algorithm is used to dynamically predict the building tilt trend, and the grouting parameters are adjusted through proportional-integral-derivative control to respond to real-time changes and generate dynamic adjustment parameters. S6: Based on dynamically adjusted parameters, the grouting equipment is controlled to perform grouting operations. Intelligent responsive grout is used and precise injection is carried out through an adaptive injection system. Priority is given to implementation in settlement areas, and grouting execution records are generated. S7: Based on grouting execution records and real-time monitoring system data, a closed-loop control algorithm is adopted to monitor grouting effect and tilt change in real time, and optimize the injection process through feedback loop, calibrate deviations in a timely manner, and generate real-time monitoring feedback. S8: Based on real-time monitoring feedback, a multi-objective optimization algorithm is used to evaluate the correction effect, and a comprehensive evaluation report is generated through Pareto frontier analysis. S9: Based on the effect evaluation report, use artificial intelligence algorithms to generate a final corrective summary and subsequent maintenance suggestions, and output visualized results through a report generation tool to ensure the continuous optimization of the solution and generate the final optimized solution.

2. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, Generating the initial dataset based on S1 involves the following steps: S101: Based on the need for high-rise building correction, a drone equipped with a lidar and a high-definition camera module is used to perform three-dimensional laser scanning of the building facade and surrounding environment, and a high-precision building geometric model is obtained through a point cloud data generation algorithm to generate the initial point cloud model of the building. S102: Based on the initial point cloud model of the building, a key point sensor network is deployed inside the building and on the foundation. The real-time deformation and tilt data of the structure are automatically collected through the wireless sensor network protocol to generate a real-time structural monitoring dataset. S103: Based on the real-time structural monitoring dataset, the foundation exploration is carried out using ground-penetrating radar and standard penetration test methods to obtain soil layer distribution, density and groundwater level data, and then the data are digitized through geological statistical analysis to generate a foundation geological attribute database. S104: Based on the initial point cloud model of the building, the real-time monitoring dataset of the structure, and the geological attribute database of the ground, a data fusion algorithm is used to perform spatiotemporal registration and integration of multi-source heterogeneous data, eliminate noise and conflicts, form a basic data package in a unified format, and generate the initial dataset.

3. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, The steps involved in generating the preprocessed data model based on S2 are as follows: S201: Based on the initial dataset, a statistical outlier detection algorithm is used to identify and remove obvious errors and noisy data points generated during the collection process, generating a cleaned dataset; S202: Based on the cleaned dataset, the principal component analysis algorithm is used to reduce the dimensionality of the high-dimensional monitoring data, extract the main feature components that best represent the building tilt and settlement trend, and generate the building settlement main feature vector. S203: Based on the main feature vector of building settlement, a convolutional neural network algorithm is used to perform deep learning on historical and real-time tilt patterns to identify settlement patterns, rates and potential risk areas, and generate a building tilt pattern identification report. S204: The preprocessed data model is generated by combining the cleaned dataset, the main feature vector of building settlement, and the building tilt pattern recognition report, and using data standardization methods to normalize all features to the same dimension and encapsulate them into standard data objects that can be called by the model.

4. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, The S3-based optimized grouting scheme includes the following steps: S301: Based on the preprocessed data model, a high-precision three-dimensional digital twin model containing the building structure, foundation soil layer and sensor points is constructed using building information modeling software to generate a building-foundation digital twin model. S302: Based on the building-foundation digital twin model, the finite element analysis algorithm is used to perform mechanical simulation under different grouting parameters, calculate stress redistribution and expected displacement field, and generate a set of grouting condition simulation results; S303: Based on the grouting condition simulation result set, a multi-objective genetic algorithm is adopted, with "maximizing the correction effect", "minimizing the material consumption" and "minimizing the construction time" as the objective functions, to perform multiple rounds of iterative optimization, search for the optimal solution set, and generate the Pareto optimal solution set; S304: Based on the Pareto optimal solution set, a decision support system is used to combine engineering experience and cost constraints to select a final implementation plan from the solution set, clarify the grouting hole location, sequence, grout ratio and pressure parameters, and generate an optimized grouting plan.

5. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, The generation of a real-time monitoring system based on S4 includes the following steps: S401: Based on the key monitoring areas identified in the optimized grouting scheme, an IoT device deployment strategy is adopted to supplement the deployment of wireless inclinometers, micromanometers, and earth pressure cell sensors to form a high-density monitoring network and generate an enhanced sensor network deployment map. S402: Based on the enhanced sensor network deployment diagram, a wireless networking protocol is used to complete the self-organizing network connection of all sensors, and a time division multiple access scheduling algorithm is used to configure the data acquisition and transmission frequency and generate a real-time data acquisition scheduling table. S403: Based on the real-time data acquisition scheduling table, cloud-based data receiving and storage services are used to build a data pipeline to receive, verify, and persistently store massive amounts of incoming real-time monitoring data, generating a cloud-based raw monitoring database. S404: Based on the cloud-based raw monitoring database, it uses a rule engine to set various alarm thresholds and configures data visualization dashboards and early warning information push mechanisms to generate a real-time monitoring system.

6. The grouting reinforcement and correction method for a high-rise building structure according to claim 1, characterized in that, The generation of dynamically adjusted parameters based on S5 includes the following steps: S501: Based on the real-time data stream pushed by the real-time monitoring system, the Kalman filter algorithm is used for data fusion and filtering to estimate the optimal value of the building tilt state, reduce the impact of measurement noise, and generate the optimal value of the building state. S502: Based on the optimal estimated value of the building status and the expected target in the optimized grouting scheme, the model predictive control algorithm is used to predict the building tilt trend in the future period and generate the tilt trend prediction value. S503: Based on the tilt trend prediction value, the proportional-integral-derivative control algorithm is used to calculate the adjustment amount required for the current grouting parameters so that the prediction value approaches the target value, and a set of grouting parameter adjustment instructions is generated. S504: Sends the grouting parameter adjustment instruction set to the grouting equipment controller, drives the intelligent grouting pump and valve actuator to make precise adjustments, realizes closed-loop adaptive control of the grouting process, and generates dynamic adjustment parameters.

7. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, The generation of grouting execution records based on S6 includes the following steps: S601: Based on dynamically adjusted parameters, a programmable logic controller is used to control the start-up, shutdown, pressure, and speed of the high-pressure grouting pump, ensuring that the output parameters are consistent with the instructions and generating pumping control signals; S602: Based on the pumping control signal, an intelligent grout delivery system is adopted to automatically mix cement, chemical grout and nano-additives according to the proportion in the optimized grouting scheme to prepare intelligent responsive grout and generate customized grout. S603: The customized grout is delivered to the grouting holes specified in the optimized grouting scheme through a distributed grouting pipeline system, and the injection status of each hole is monitored in real time using flow meters and pressure sensors to generate real-time data on hole injection. S604: Based on pumping control signals and real-time data of hole injection, a data log recording algorithm is adopted, using timestamps as indexes, to fully record the grouting parameters, equipment status and injection volume at each moment, and generate grouting execution records.

8. The grouting reinforcement and correction method for high-rise building structures according to claim 1, characterized in that, Generating real-time monitoring feedback based on S7 includes the following steps: S701: Based on the continuous data stream of grouting execution records and real-time monitoring system, a time series comparison algorithm is used to compare the actual settlement / lifting data after grouting with the tilt trend prediction value in real time, and generate a real-time comparison report of grouting effect. S702: Based on the real-time comparison report of grouting effect, the deviation analysis algorithm is used to calculate the error between the actual effect and the expected effect, locate the grouting area with poor effect or abnormality, and generate a grouting deviation analysis diagram. S703: Feeds back the grouting deviation analysis diagram to the model predictive control algorithm of S5 as the input for a new round of predictive control, corrects the predictive model parameters in real time, optimizes the accuracy of subsequent control commands, and generates model parameter correction amounts. S704: Based on the model parameter correction amount, a control loop iterative mechanism is adopted to dynamically update the control strategy and generate decision suggestions on whether to supplement grouting or adjust the focus, and generate real-time monitoring feedback.

9. The grouting reinforcement and correction method for a high-rise building structure according to claim 1, characterized in that, The S8-based performance evaluation report includes the following steps: S801: Based on real-time monitoring feedback and grouting execution records, a data aggregation algorithm is used to extract key performance indicators such as the final inclination, total grout usage, and total man-hours after the entire construction cycle, and generate a set of performance indicators for the correction project. S802: Based on the performance index set of the correction project, a multi-attribute decision analysis method is adopted, and a comprehensive evaluation system is constructed by combining the dimensions of economic efficiency, environmental impact and social benefits, and a multi-dimensional evaluation matrix is ​​generated. S803: Based on a multi-dimensional evaluation matrix, a particle swarm optimization algorithm is used for non-dominated sorting to find a set of solutions that perform well under multiple objectives, forming a Pareto front and generating a comprehensive evaluation Pareto front. S804: Based on a comprehensive assessment of the Pareto frontier, a report generation engine is used to automatically generate a structured assessment report, which includes key findings, scores for each indicator, cost-benefit analysis, and improvement suggestions, thus generating an effectiveness assessment report.

10. The grouting reinforcement and correction method for a high-rise building structure according to claim 1, characterized in that, The final optimized solution based on S9 includes the following steps: S901: Based on the effectiveness evaluation report, use natural language processing technology to extract key conclusions, successful experiences and existing problems from the report, and generate a key information extraction summary; S902: Based on key information extraction, a summary is extracted, and reinforcement learning is used to simulate the long-term effects of different subsequent maintenance strategies. The agent is trained to learn the optimal long-term maintenance scheme and generate simulation results of long-term maintenance strategies. S903: Based on the simulation results of long-term maintenance strategy, the case reasoning method is used to store all data, models and results of this project into the knowledge base to provide reference for similar projects in the future and generate grouting correction knowledge base update records; S904: Integrates key information extraction summaries, long-term maintenance strategy simulation results, and grouting correction knowledge base update records, and uses intelligent document generation technology to output the final summary report, operation and maintenance manual, and sustainable optimization suggestions, generating the final optimization plan.

Images