Complex geology pile releasing construction method and system based on BIM (Building Information Modeling) technology

Abstract

The invention relates to the technical field of complex geology pile releasing construction methods based on the BIM technology, and discloses a construction method named as a complex geology pile releasing construction method based on the BIM technology. The method aims at solving the problems of design and construction disjunction, insufficient risk pre-control, low resource utilization rate and the like caused by geological information dimension reduction processing and construction process static management in the prior art. The method is characterized by comprising the following steps: constructing a multi-source heterogeneous geological information full-parameterization three-dimensional attribute body model, and carrying out integrated simulation optimization design on a single-group pile cap based on the model; and sensing construction parameters in real time to dynamically correct the geologic model, so as to generate a self-adaptive construction control instruction and high-precision lofting data. By the adoption of the technical scheme, the safety, economical efficiency and environment friendliness of pile foundation engineering can be remarkably improved.

Description

Technical Field

[0001] This invention belongs to the technical field of civil engineering construction methods, and specifically relates to a complex geological pile laying construction method based on BIM technology. Background Technology

[0002] With the widespread application of BIM (Building Information Modeling) technology in the field of engineering construction, pile foundation construction under complex geological conditions is gradually shifting from traditional experience-driven methods to digital and intelligent approaches. However, existing BIM-based pile foundation construction methods still have shortcomings in geological information processing, dynamic management of the construction process, and collaborative operation capabilities, making it difficult to meet the needs of refined management of the entire life cycle of pile foundation projects under complex geological conditions.

[0003] A search revealed a patent, CN119227207B, which discloses an automatic generation method for municipal road construction layout information based on a BIM model, published on February 11, 2025. This patent improves the computational efficiency and ease of interpretation of construction layout information by transforming the BIM model coordinate system, adding layout attribute fields, and batch exporting construction layout information reports. It combines 3D views to generate construction layout information reports that integrate models, views, and data tables, effectively enhancing the computational efficiency and ease of interpretation of construction layout information. However, this technical solution primarily targets municipal road engineering, focusing on the automated generation of layout information. It lacks support for high-fidelity modeling and dynamic updating of geological information under complex geological conditions, failing to achieve real-time linkage between geological information and the construction process, thus limiting its applicability in complex geological pile-laying construction. Furthermore, this solution does not address risk pre-control and resource optimization during construction, making it difficult to cope with the uncertainties under complex geological conditions.

[0004] A search revealed a method for constructing ultra-deep test piles in rock strata, published on August 17, 2021, with publication number CN111519613B. This patent addresses the challenges of drilling in complex rock strata and low accuracy in controlling over-pouring elevation by establishing a BIM model of the rock surface through advanced drilling exploration, combined with multi-stage step drilling, dual-control verticality monitoring (both internal and external), and underwater concrete pouring monitoring. However, this technical solution is only applicable to specific rock strata test pile construction scenarios. Geological information processing is limited to static modeling, lacking a real-time perception and feedback mechanism for dynamic geological changes, making it difficult to adapt to the dynamic changes in the construction environment under complex geological conditions. Furthermore, this solution does not establish a collaborative operation system throughout the entire lifecycle of the pile foundation project, resulting in a disconnect between information transmission between the design and construction stages, leading to low resource utilization and limited construction efficiency.

[0005] The aforementioned problems indicate that existing BIM-based pile foundation construction methods still have significant shortcomings in high-fidelity dynamic modeling of geological information, real-time collaborative management of the construction process, and risk prevention and resource optimization throughout the entire lifecycle. Therefore, this invention provides a BIM-based pile laying construction method and system for complex geological conditions. The aim is to establish a dynamic closed-loop collaborative operation system based on a high-fidelity geological digital twin model that spans the entire lifecycle of pile foundation engineering. This system will improve the integration of design and construction in complex geological conditions, enhance risk prevention and control capabilities, and increase resource utilization efficiency, thereby meeting the demands of modern engineering construction for intelligent and refined construction. Summary of the Invention

[0006] The purpose of this invention is to overcome the technical problems in the prior art, such as the disconnect between design and construction, insufficient risk prevention and control capabilities, and low resource utilization caused by the dimensionality reduction processing of geological information and static management of the construction process. It provides a complex geological pile laying construction method and system based on BIM technology, aiming to establish a dynamic closed-loop collaborative operation system based on a high-fidelity geological digital twin model that runs through the entire life cycle of pile foundation engineering.

[0007] To achieve the above-mentioned objectives, this invention provides a method for complex geological pile laying construction based on BIM technology, characterized by the following steps: Step 1: Construct a fully parameterized 3D geological attribute volume model integrating multi-source heterogeneous geological information. This step is not a simple 3D geometric connection of discrete borehole data, but rather a systematic data fusion and spatial interpolation process to generate a digital twin with voxels as the basic unit, capable of characterizing the physical and mechanical properties of the continuous underground medium. Specifically, firstly, through a standardized data acquisition interface, geological information from at least the following sources is collected: First, existing borehole exploration data of the site, including the boundaries of soil and rock layers, soil and rock types, standard penetration test blow counts, and geotechnical test parameters at different depths of each borehole; second, wide-area geophysical exploration data, specifically 3D data fields of apparent resistivity or shear wave velocity covering the entire construction area obtained by high-density resistivity or surface wave exploration methods; third, regional geological history data and engineering analogy data. Subsequently, a geostatistical modeling engine was launched, employing a co-kriging interpolation algorithm. High-precision borehole location data served as hard data constraints, while comprehensive geophysical exploration data provided soft data trend guidance. This process involved three-dimensional meshing of the entire target geological space, forming a voxel set. For each voxel, the engine calculated and assigned a set of probabilistic physical and mechanical parameters, including at least: soil / rock type and its confidence level, water content, void ratio, compression modulus, internal friction angle, cohesion, permeability coefficient, and shear wave velocity. The resulting three-dimensional geological attribute model is a digital geological entity stored in a database. Each voxel carries a complete set of parameters, accurately reflecting the spatial gradation of lithology, the morphology of weak interlayers, the three-dimensional boundaries of water-rich areas, and the uneven distribution of soil / rock mechanical properties.

[0008] Step Two: Perform integrated performance simulation and multi-objective collaborative optimization design of single pile-pile group-pile cap based on the aforementioned 3D geological property volume model. This step aims to transform pile foundation design from traditional independent calculations based on the assumption of two-dimensional homogeneous soil layers to integrated system performance optimization in a real three-dimensional heterogeneous geological environment. Specifically, firstly, the Building Information Model (BIM) of the superstructure and the aforementioned 3D geological property volume model are precisely registered and coupled in a unified coordinate system to form a complete calculation model including superstructure loads, pile foundation structure, and geological foundation. Subsequently, an embedded finite element analysis (FEA) solver is used to perform loading simulation on the coupled model. This simulation process is based on the voxel mesh of the aforementioned 3D geological property volume model, directly inputting the physical and mechanical parameters of each voxel as the material properties of the finite element element, thereby enabling accurate simulation of the axial force transmission, lateral friction distribution, pile end reaction force, and pile settlement, differential settlement, and group pile effect caused by pile-soil interaction for each pile under complex geological conditions. Based on this, a multi-objective optimization function is defined, whose objective variables include at least: total project cost, overall structural safety factor, and maximum differential settlement value. The constraints of the function include various bearing capacity and deformation limits required by national standards. An iterative optimization calculation is performed using a genetic algorithm or particle swarm optimization algorithm, with the pile length, pile diameter, pile type (e.g., precast pile, cast-in-place pile, MJS method pile), pile reinforcement ratio, and fine-tuning coordinates of the pile location within the allowable range as decision variables. Each iteration of the algorithm is evaluated using a FEA solver until a pile foundation layout and parameter scheme that satisfies all constraints and achieves a Pareto optimal solution set for the multi-objective optimization function is found. This scheme specifies optimized, personalized design parameters for each pile and generates a pile foundation BIM model containing the 3D model, precise coordinates, and material specifications of all piles.

[0009] Step 3: Implement dynamic correction and assimilation of the geological property model based on real-time parameter perception during construction. The core of this step lies in establishing a real-time data feedback loop from the construction site to the digital model, transforming the construction process itself into a dynamic exploration process for continuous verification and refinement of the geological model. Specifically, a data acquisition sensor array is deployed on the key execution components of the pile foundation construction equipment (such as rotary drilling rigs and static pressure pile drivers). This array includes at least: a torque-speed sensor for measuring drill bit torque and rotational speed; a displacement-velocity sensor for measuring drilling depth and rate; a pressure-flow sensor for measuring the inlet and outlet pressure and flow rate of the mud circulation system; and an online spectral analyzer or image recognition module for analyzing the lithology of drill cuttings. During construction, the sensor array continuously collects construction parameters at a preset sampling frequency (e.g., once every 10 cm of drilling). The collected raw data is preprocessed and packaged through an edge computing gateway and uploaded in real-time to a central server via an industrial wireless network (such as 5G-U). The server runs a geological parameter inversion analysis model. Based on the theoretical relationship between rock and soil mechanics and borehole engineering, this model inversely analyzes the input construction parameters (such as torque and drilling speed) into equivalent geological parameters of the soil and rock strata at the drill bit location (such as drill resistance and formation hardness index). Subsequently, a data assimilation module is activated, employing Kalman filtering or ensemble Kalman filtering algorithms to update and correct the original probabilistic parameters of the corresponding voxels in the three-dimensional geological attribute volume model, using the inverted formation parameters as new observations. This process is continuous; that is, after each pile or section of a pile is completed, the geological model is updated by the actual exposed formation conditions, gradually reducing uncertainty and continuously improving accuracy.

[0010] Step Four: Generate and issue adaptive construction control instructions and high-precision layout data based on the dynamically corrected model. This step is the execution link of the closed-loop control, which transforms the updated high-precision model information into precise guidance and risk avoidance instructions for on-site construction. Specifically, on the one hand, the precise three-dimensional spatial coordinates of each pile, optimized and solidified in the pile foundation BIM model in Step Two, are directly output as layout data files recognizable by construction surveying robots or construction equipment with RTK-GPS positioning capabilities. On-site layout operations are transformed from traditional "point-line" control to precise spatial positioning based on the three-dimensional model, with positioning errors controlled at the centimeter level. On the other hand, a feedforward risk prediction and control module, based on the latest corrected geological attribute volume model, performs advanced simulation of the pile segment to be constructed. For example, it simulates the weak interlayers, high-pressure water-rich sand layers, or karst caves that may be encountered during the next meter of drilling, and quantitatively assesses the probability of accidents such as borehole collapse, water inrush, and stuck drill. When the predicted risk probability exceeds the preset safety threshold, the module automatically generates adaptive construction control instructions. The specific instructions include: adjusting the drilling pressure and rotation speed of the rotary drilling rig to adapt to changes in the hardness of the formation; adjusting the output flow rate and pressure of the mud pump to dynamically balance the water and soil pressure on the borehole wall; or automatically adjusting the mud mix ratio (such as increasing the bentonite content to improve wall protection performance) before entering specific high-risk formations. These instructions are sent directly to the PLC (Programmable Logic Controller) of the construction equipment in the form of digital signals to achieve semi-automatic or fully automatic closed-loop adjustment of construction process parameters.

[0011] Furthermore, as a preferred embodiment of the method of the present invention, when dealing with specific water-rich silty sand strata or scenarios requiring deep foundation pit excavation next to existing buildings, the multi-objective collaborative optimization design step in step two will automatically trigger and select a composite support and water-stopping structure scheme formed by combining MJS (Metro Jet System) piles and bored interlocking piles. Specifically, the Pile Foundation Intelligent Design Unit, based on the three-dimensional spatial distribution of the high-permeability, low-cohesion silty sand layer identified in the geological model, first designs a bored interlocking pile water-stopping curtain composed of C30 plain concrete piles with a diameter of 800mm. The piles are semi-interlocked, with an interlocking depth of not less than 250mm, ensuring that the bottom of the curtain is embedded at least 1.5 meters into the underlying relatively impermeable layer. Subsequently, inside this curtain, MJS piles are designed to reinforce the soil at the foundation pit excavation surface. The MJS method piles employ horizontal jet grouting. The jetting radius, pressure, duration, and lifting speed are all precisely calculated using optimization algorithms based on the physical and mechanical parameters (internal friction angle, cohesion) of the soil to be reinforced. This ensures that the resulting consolidated body achieves the required unconfined compressive strength (e.g., not less than 1.5 MPa). This combined approach utilizes a BIM model for integrated design and clash detection, ensuring spatial synergy and functional complementarity between the two pile types. This enables deterministic control of groundwater and soil deformation under complex geological conditions.

[0012] Furthermore, to address the challenges and high risks associated with installing auxiliary facilities such as water intake and cofferdams during bridge pier pile foundation construction in underwater environments like fast-flowing riverbeds, this invention integrates a targeted underwater construction adaptive control module in step four. This module first calculates the flow field distribution in the fast-flowing area using fluid dynamics simulation (CFD) software, based on the riverbed topography data and hydrological survey data contained in the three-dimensional geological property model. Based on the simulation results, an optimized layout scheme for a flow guidance system composed of prefabricated steel structure water-retaining plates is designed. This scheme aims to reduce the water flow velocity at the pile location to a safe operating window (e.g., below 0.5 m / s). During construction, the water-retaining plates are first precisely installed. Subsequently, an underwater overburden blasting or bedrock drilling operation is performed using a "three-pipe two-drill method." Specifically, the "three-pipe two-drill method" refers to a combined drilling platform consisting of a central positioning steel pipe and two symmetrically arranged working steel pipes. The central positioning steel pipe is first precisely sunk to the design elevation using a vibratory pile driving method, serving as the benchmark for subsequent operations. Two down-the-hole drilling rigs were lowered into the two working steel pipes, and parameters such as torque and drilling speed during the drilling process were monitored in real time by the system in step three. When the monitoring data showed that hard rock layers were encountered, the adaptive control system instructed the drilling rigs to stop drilling and dropped precisely measured underwater emulsion explosives for micro-blasting to break the rock or form anchoring grooves for pile embedding. The entire process was monitored and guided by three-dimensional visualization using a BIM model, ensuring construction accuracy and safety under complex hydrogeological conditions.

[0013] Furthermore, in engineering scenarios requiring the construction of vertical seepage barriers to prevent the spread of pollution sources or to prevent lateral seepage from foundation pits, the method of this invention implements an intelligent seepage barrier construction technology based on modified kaolin grouting material in steps two and four. Specifically, in the design stage of step two, based on the three-dimensional distribution map of permeability coefficient in the three-dimensional geological property model, the spatial range, depth, and thickness of the seepage barrier to be constructed are precisely delineated. Simultaneously, based on the chemical composition, pore structure, and hydraulic gradient of the target soil layer, the optimal grouting material formula is determined through a material ratio optimization algorithm. This formula uses kaolin as the base material (with a particle size D50 controlled at 2-5 micrometers), and incorporates a certain proportion (e.g., 5%-10% by weight) of sodium bentonite to enhance its expansion and water-stopping performance, and 0.5%-1.5% by weight of sodium carboxymethyl cellulose (CMC) as a dispersant and water-retaining agent. In the construction stage of step four, grouting operations are carried out using a constant pressure, quantitative, and timed method. The outlet pressure and injection flow rate of the grouting pump are controlled in a closed loop by an adaptive control system. Based on the real-time grouting pressure fed back by the pressure sensor in step three, the control system compares it with the theoretical soil and water pressure at the corresponding location in the geological model, dynamically adjusting the injection rate to prevent fracturing grouting or insufficient compaction. The resulting impermeable wall can stably achieve a permeability coefficient of 10. -8On the order of cm / s, a low-permeability barrier was constructed that is reliable and highly compatible with the surrounding geological environment.

[0014] Furthermore, the present invention also provides a complex geological pile-laying construction system based on BIM technology, which is configured to perform any of the above methods, characterized in that it includes: A geological information fusion and modeling unit integrates a standardized data acquisition interface, a co-kriging geostatistical modeling engine, and a three-dimensional voxel model database. The unit is configured to receive multi-source heterogeneous geological information and generate the fully parameterized three-dimensional geological attribute volume model.

[0015] A pile foundation performance simulation and optimization design unit includes a high-performance computing server cluster on the hardware side and a finite element analysis solver and a multi-objective genetic optimization algorithm program that couples BIM and geological models on the software side. The unit communicates bidirectionally with the geological information fusion and modeling unit and is configured to perform integrated performance simulation of single pile-pile group-cap and output the optimized pile foundation BIM model.

[0016] A construction process dynamic monitoring unit includes a sensor array deployed on the construction equipment, an edge computing gateway, and a geological parameter inversion analysis module. The unit is configured to collect, process, and upload construction process parameters in real time, and invert them into stratigraphic equivalent geological parameters.

[0017] An adaptive closed-loop control unit integrates a Kalman filter data assimilation module, a feedforward risk prediction module, and a PLC control instruction generation module. This unit interacts with the geological information fusion and modeling unit, the pile foundation performance simulation and optimization design unit, and the construction process dynamic monitoring unit, respectively. It is configured to dynamically correct the geological model, generate high-precision layout data and adaptive construction control instructions based on the corrected model, and send the instructions to the execution mechanism of the construction equipment.

[0018] A waste mud recycling module, as a supplement to the entire closed-loop system, includes online monitoring sensors for mud properties (such as densitometers and viscometers), an automatic flocculant and solidifying agent dosing system consisting of metering pumps and mixers, and a solidified soil mechanical property testing module. This module monitors the physicochemical properties of the waste mud generated during construction in real time, automatically calculates and adds the optimal dosage of flocculant and solidifying agent (such as cement and lime), transforming the waste mud into solidified soil with specific bearing capacity (e.g., unconfined compressive strength not less than 0.5 MPa). Its physical and mechanical parameter data is then written back into the BIM model for unified scheduling and management as usable materials for subsequent roadbed backfilling or site hardening, thereby achieving closed-loop utilization of construction waste.

[0019] The geographically dispersed units in the system are interconnected through a low-latency, high-bandwidth communication network based on fiber optics and a 5G industrial private network, ensuring the real-time performance and reliability of the entire information flow from data acquisition, model updates, decision generation to on-site control. Through the above method and system, this invention integrates the originally separate and static design and construction processes into an intelligent collaborative operation system based on a digital twin model, possessing self-sensing, self-learning, self-decision-making, and self-execution capabilities. This fundamentally solves the deep-seated contradictions faced by pile foundation engineering under complex geological conditions, significantly improving the safety, economy, and environmental friendliness of the project. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the method of the present invention. Figure 2 This is a structural block diagram of the system of the present invention.

[0021] The attached diagram is labeled as follows: 10, Geological Information Fusion and Modeling Unit; 20, Pile Foundation Performance Simulation and Optimization Design Unit; 30, Construction Process Dynamic Monitoring Unit; 40, Adaptive Closed-Loop Control Unit; 50, Waste Slurry Resource Utilization Module. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Those skilled in the art should recognize that various modifications, equivalent substitutions, and improvements to the technical solutions of this invention can be made without inventive effort and are all within the scope of protection of this invention.

[0023] To systematically address the inherent limitations of existing technologies in geological information processing, pile foundation design, construction control, and risk management, this invention proposes a complex geological pile laying construction method and its supporting system based on BIM technology. This method aims to construct a high-fidelity geological digital twin and establish a dynamic closed loop around this twin, from design optimization to adaptive construction control, thereby achieving intelligent and refined management of the entire lifecycle of pile foundation engineering. (Refer to...) Figure 1 The flowchart shown clearly illustrates the core technology of this invention. It is not only a linear sequence of steps, but also a cyclical improvement system that includes data feedback and model iteration.

[0024] In a specific implementation scenario, the execution of the method of this invention begins with step one, namely, constructing a fully parameterized three-dimensional geological attribute model that integrates multi-source heterogeneous geological information. The goal of this step is not to generate a simple three-dimensional geometric visualization model, but to create a computable digital twin capable of accurately representing the continuous changes in the physical and mechanical properties of underground soil and rock media in three-dimensional space. This process is first executed through a geological information fusion and modeling unit 10. This unit 10 is equipped with a standardized data acquisition interface, capable of seamlessly accessing and parsing geological exploration data from various formats and sources. Specifically, the acquired data includes at least three levels: firstly, high-precision borehole exploration data obtained during the site detailed exploration phase. These data are stored in structured tables, with each record corresponding to a sampling point in a borehole. Fields include borehole number, three-dimensional coordinates (X, Y, Z), soil and rock layer description, soil and rock category code (according to the "Code for Geotechnical Investigation" GB 50021), standard penetration test (SPT) measured blow count N value, and a series of key parameters obtained through laboratory geotechnical tests, such as water content, void ratio, liquid limit, plastic limit, compression modulus, internal friction angle, and cohesion. Secondly, there is wide-area geophysical exploration data covering the entire construction site. For example, a GeodeCMS manufactured by Geometrics Corporation in the United States can be used. TM Ultra-high density resistivity meters, using a Wenner array to deploy survey lines with a spacing of 20 meters and an electrode spacing gradually increasing from 5 meters to 80 meters, acquire a three-dimensional apparent resistivity data cube covering the target depth range (e.g., 50 meters underground). Alternatively, the surface wave exploration method (MASW) is used, recording Rayleigh waves generated by active sources (such as hammer strikes) using multichannel seismographs to invert a three-dimensional shear wave velocity (Vs) data field covering the entire field. Although these geophysical data have lower resolution than borehole data, they provide trend information reflecting stratigraphic structure and lithological changes across the entire area. Thirdly, there are regional geological history data and similar engineering experience data, such as descriptions of regional tectonics and paleochannel distribution in local geological records, as well as settlement observation records and post-pile foundation evaluation reports from completed projects on adjacent sites.

[0025] After data aggregation, the geostatistical modeling engine within the geological information fusion and modeling unit 10 is activated. The core algorithm of this engine is the co-kriging spatial interpolation algorithm. Within this algorithm framework, high-precision but sparsely distributed borehole location data are defined as "hard data" or main variables, serving as strong constraints on the interpolation results; while geophysical exploration data with full coverage but relatively low precision (such as apparent resistivity or shear wave velocity) are defined as "soft data" or covariates, providing trend guidance on spatial variability for the interpolation process. For example, soil density has a strong positive correlation with shear wave velocity and resistivity. Utilizing this correlation, the co-kriging algorithm can more reasonably infer the gradual changes in soil parameters in the unsampled area between two boreholes based on the changing trends of geophysical data. The modeling engine first divides the entire target geological space (e.g., a cuboid region 500 meters long, 300 meters wide, and 60 meters deep) into a 3D mesh, forming a set of numerous voxels. The size of each voxel is set according to the required precision, for example, 1 meter × 1 meter × 0.5 meters. Then, for each voxel, the engine calculates and assigns a set of probabilistic physical and mechanical parameters based on its surrounding hard and soft data points by solving the Kriging equations. These attributes include not only the most likely soil or rock type and its confidence level (e.g., 75% probability of silty clay and 25% probability of silty clay), but also the expected value and variance of parameters such as water content, void ratio, compression modulus Es, internal friction angle φ, cohesion c, permeability coefficient k, and shear wave velocity Vs. Finally, a large 3D geological attribute volume model is constructed and stored in a PostgreSQL database configured with PostGIS spatial extensions. Each voxel in this model is an information carrier, and together they constitute a digital geological entity that can accurately reflect the spatial heterogeneity of strata lithology, the occurrence morphology of weak interlayers, the three-dimensional boundaries of lenses, the distribution of water-rich areas, and the heterogeneity of rock and soil mechanical properties, laying a solid data foundation for subsequent refined design and simulation.

[0026] Next, the process moves to step two, executing integrated performance simulation and multi-objective collaborative optimization design of single pile-pile group-foundation based on the aforementioned three-dimensional geological attribute model. This step is implemented by a pile foundation performance simulation and optimization design unit 20, which is typically deployed on a high-performance computing server cluster. Its core idea is to completely abandon the traditional design practice of simplifying the foundation to a horizontal homogeneous soil layer, and instead perform coupled analysis and performance optimization of the entire "superstructure-foundation-pile group-foundation" system in a highly realistic three-dimensional heterogeneous geological environment. Specifically, firstly, the superstructure BIM model completed by the architect is imported into the system through a standardized data interface (such as Industry Foundation Classes, IFC 4.0 standard). Simultaneously, the three-dimensional geological attribute model generated in step one is retrieved from the database. Under a unified engineering coordinate system (e.g., an independent construction coordinate system established based on the National 2000 coordinate system), the two are precisely spatially registered and data coupled through shared reference control points. At this point, a complete calculation model containing superstructure load information (obtained through the BIM model), pile foundation structure (initial layout), and detailed geological foundation is completed.

[0027] Subsequently, the finite element analysis (FEA) solver integrated within unit 20 (e.g., commercial software ABAQUS or PLAXIS 3D) is invoked to perform a loading simulation analysis on the coupled model. A key feature of this solver is its ability to directly read and process voxelized geological models. During the finite element mesh generation phase, the mesh for the geological portion is consistent with the voxel mesh or directly converted from the voxel mesh. The material properties of each soil element (e.g., using C3D8R eight-node linear reduced integral elements) are dynamically assigned directly from the physical and mechanical parameters (such as compression modulus, internal friction angle, cohesion, etc.) carried by its corresponding voxel. The material constitutive model can be selected according to the soil type; for example, a modified Cambridge model is used for cohesive soils, and a Mohr-Coulomb or more advanced hardened soil model is used for sandy soils. By applying dead loads, live loads, and loads such as wind and earthquakes transmitted from the superstructure, the FEA solver can accurately simulate the axial force transmission curve of each pile, the distribution law of the pile body's lateral friction, the performance of the pile end resistance, and more importantly, the absolute settlement of the pile body, the differential settlement of the pile foundation as a whole, and the complex group pile effect caused by the interaction between the pile and the soil under complex heterogeneous geological conditions.

[0028] With accurate simulation capabilities in place, a multi-objective collaborative optimization design process was initiated. First, a comprehensive multi-objective optimization function was defined, aiming to simultaneously seek the Pareto optimal solution for multiple mutually constraining objectives. Its objective variables typically include: F1 = total project cost (directly related to pile length, pile diameter, concrete grade, reinforcement ratio, etc.), F2 = overall structural safety factor (to be maximized), and F3 = maximum differential settlement value (to be minimized). Simultaneously, a series of strict constraints were set, such as: the vertical bearing capacity of a single pile must meet the code requirements; the strength of the pile material must not exceed limits; the punching shear and shear resistance of the pile cap must meet requirements; and the maximum settlement and differential settlement must not exceed the code limits. The decision variables for optimization are the design parameters that can be adjusted within the allowable range of the project. These include: the pile length, pile diameter, pile type (e.g., under the same pile cap, precast square piles, bored piles, or MJS method piles can be used in combination depending on the geological conditions at different locations), the concrete grade and reinforcement ratio of the pile body, and even the allowable fine-tuning coordinates of the pile position within the pile cap (e.g., ±100mm). The optimization algorithm can be a robust heuristic algorithm suitable for handling complex nonlinear problems, such as the Non-Dominated Sorting Genetic Algorithm (NSGA-II) or the Multi-Objective Particle Swarm Optimization Algorithm (MOPSO). Each iteration of the algorithm generates a new set of pile foundation design schemes (an individual or particle), which is immediately fed into the FEA solver for performance evaluation, obtaining the corresponding objective function values ​​such as cost, safety factor, and differential settlement. Through hundreds or thousands of generations of evolution (or iteration), the algorithm eventually converges to a Pareto Front, which represents a set of "non-dominated" optimal solutions. Project decision-makers can select the final solution from this solution set that best meets the project requirements, based on their different priorities regarding cost, safety, and deformation control. This solution assigns optimized, highly personalized design parameters to each pile on the site. Finally, the pile foundation performance simulation and optimization design unit 20 outputs a fully optimized pile foundation BIM model that includes all piles (including their precise three-dimensional geometric model, spatial coordinates, material specifications, reinforcement information, etc.), serving as a precise basis for subsequent construction.

[0029] Once the construction phase begins, the innovation of this invention lies in step three: implementing dynamic correction and assimilation of the geological attribute model based on real-time parameter perception during the construction process. The essence of this step is to establish a real-time data feedback loop from the physical construction site to the digital twin model, cleverly transforming the pile foundation construction process itself into a dynamic "secondary exploration" process for continuous verification, correction, and refinement of the geological model. To this end, a robust and durable data acquisition sensor array is integrated into the key execution components of the pile foundation construction equipment (such as the Bauer BG series or Sany Heavy Industry SR series rotary drilling rigs). This sensor array serves as the physical front end of a dynamic monitoring unit 30 for the construction process. Specifically, it includes: a high-precision torque-speed sensor mounted on the power output shaft of the drill bit drive head, used to measure in real time the torque required for the drill bit to break up rock and soil and the rotational speed of the drill bit; a high-precision displacement-velocity sensor (such as a laser rangefinder or photoelectric encoder) mounted on the hoisting system or the vertical guide rail of the power head, used to accurately record the real-time drilling depth and drilling rate of the drill bit; industrial-grade pressure-flow sensors installed on the inlet and outlet pipes of the mud circulation system, used to monitor changes in the circulation pressure and flow rate of the wall-protecting mud, which can indirectly reflect the stability of the borehole wall; and, on the drill cuttings conveyor belt discharged from the drilling rig, an online spectral analyzer (such as using near-infrared spectroscopy technology) or a machine vision-based image recognition module can be integrated for real-time analysis of the mineral composition, water content, and particle morphology of the drill cuttings, thereby directly identifying the rock and soil type.

[0030] During construction, this sensor array continuously collects massive amounts of construction process parameters at a preset high sampling frequency (e.g., data is collected every 10 centimeters of drilling or every 5 seconds). These raw data streams are first sent to an edge computing gateway deployed at the construction site. This gateway performs preliminary data cleaning (noise removal), aggregation (averaging by depth or time), and formatting and packaging, then uploads the data in real-time via a low-latency, high-bandwidth industrial wireless network (such as a private network based on 5G-U technology) to a central server deployed in the cloud or at the project headquarters. A core geological parameter inversion analysis model runs on the server. This model is based on profound geotechnical mechanics and drilling engineering theories, such as the Warren formula or its improved forms, establishing mathematical or machine learning relationships between construction parameters such as drill bit torque, drill pressure, and rotation speed, and equivalent mechanical parameters of the soil strata at the drill bit location (such as the drill resistance of the foundation soil, the formation hardness index, and even the equivalent internal friction angle and cohesion).

[0031] The inverted stratigraphic parameters are considered as new, real-time "observations" of the geological conditions. Subsequently, a data assimilation module is activated, deployed within the adaptive closed-loop control unit 40. This module employs advanced filtering algorithms, such as Extended Kalman Filter (EKF) or Ensemble Kalman Filter (EnKF), to fuse these new observations with the original probabilistic parameters of the corresponding voxels in the 3D geological attribute volume model. The logic of Kalman filtering here is as follows: the original parameters of the model (expected value and variance) are used as the "predicted state," and the inverted parameters are used as the "observed state." The algorithm corrects the predicted state by calculating the Kalman gain, resulting in an "updated state" with smaller variance and closer to the true value. This process is continuous and rolling. That is, whenever a section of a pile is completed, the revealed true stratigraphic information is used to "refresh" the geological model within its influence range, gradually reducing the uncertainty of the model as construction progresses, and continuously improving accuracy and fidelity. For example, when a drilling rig in an area that was originally predicted by the model to be a medium-dense sand layer, it returns continuous high torque and low drilling speed. The inversion model judges that this area is a high-dense cemented gravel layer. The data assimilation module will then correspondingly increase the expected values ​​of parameters such as the compressive modulus and internal friction angle of the voxels in this area and reduce their variance.

[0032] The value of the dynamically corrected high-precision model is ultimately reflected in step four, which involves generating and issuing adaptive construction control commands and high-precision layout data based on this model. This is the execution stage of the entire closed-loop control system, and its core task is to transform the updated and more reliable model information into precise guidance for on-site construction operations and proactive avoidance of potential risks. This function is mainly implemented by the adaptive closed-loop control unit 40. On the one hand, this unit 40 directly extracts the precise three-dimensional spatial coordinates (X,Y,Z) of the design pile center point of each pile from the data optimized in step two and stored in the pile foundation BIM model, and converts it into layout data files (such as SDR or LandXML format) that can be directly recognized and imported by specific brand construction surveying robots (such as Trimble SX12 or Leica TS16) or construction equipment with RTK-GNSS high-precision positioning function (such as rotary drilling rigs with integrated dual-antenna RTK). The on-site layout work has thus been upgraded from the traditional two-dimensional positioning of "point-line" using measuring tapes and theodolites to precise positioning guided by spatial point clouds based on three-dimensional BIM models. The positioning errors in plane and elevation can be stably controlled within the centimeter level of ±10 millimeters, ensuring the geometric accuracy of pile foundation construction from the source.

[0033] On the other hand, a more crucial feedforward risk prediction and control module operates within this unit 40. Based on the latest revised geological property model, this module performs advanced simulation calculations for the next pile to be constructed or the next drilling segment of the current pile (e.g., the next 2 meters). For example, it can simulate whether the drill bit will suddenly encounter a newly revised weak silt interlayer in the model, enter a high-pressure confined water-rich sand layer, or an unexplored karst cave during the next meter of drilling. Through these advanced simulations, the module can quantitatively assess the probability of construction accidents such as borehole collapse, water inrush, stuck drill, and drill pipe deviation under the current construction process parameters. When the predicted risk probability exceeds the project's preset safety threshold (e.g., borehole collapse risk probability greater than 15%), the module will automatically trigger and generate a series of adaptive construction control commands. These instructions are specific and executable. For example, 0.5 meters before entering a weak interlayer, the rotary drilling rig is instructed to automatically reduce the drilling pressure from 100kN to 60kN and increase the rotation speed from 30rpm to 45rpm to pass through quickly with "fast rotation and light pressure" to prevent excessive disturbance to the borehole wall. 1 meter before encountering a confined aquifer, the mud preparation system is instructed to automatically increase the proportion of high-molecular polymers in the circulating mud and increase the output pressure and flow rate of the mud pump by 15% to dynamically balance the water and soil pressure outside the borehole wall and actively protect the wall. Alternatively, when it is predicted that the rig may enter a karst cave area, the system will issue a warning to the operator and suggest adopting a low-impact, low-disturbance drilling mode, while preparing sealing and grouting materials. These digital control commands are sent directly to the PLC (Programmable Logic Controller, such as Siemens S7-1500 series) at the bottom layer of the construction equipment via industrial Ethernet or fieldbus (such as CAN bus). The PLC precisely controls the actuators of the hydraulic system, power system and mud system, thereby realizing semi-automatic or even fully automated closed-loop adjustment of construction process parameters, turning passive accident handling into proactive risk avoidance.

[0034] As a preferred embodiment of the method of the present invention, in dealing with specific complex geological conditions, such as deep foundation pit projects in urban centers where highly permeable, water-rich fine sand strata are widely distributed, and the foundation pit is adjacent to existing subway tunnels or historical buildings, the requirements for deformation control are extremely stringent. In this scenario, the pile foundation performance simulation and optimization design unit 20 in step two will automatically trigger and optimize a composite support and water-stopping structure scheme formed by combining MJS (Metro Jet System) piles and bored interlocking piles based on the three-dimensional spatial distribution characteristics (thickness, burial depth, continuity) of the highly permeable, low-cohesion silty sand layer identified in the geological model. In specific implementation, the optimization design unit first designs a bored interlocking pile water-stopping curtain along the excavation boundary of the foundation pit, consisting of plain concrete piles (first-stage piles) with a diameter of 800mm and a concrete strength grade of C30 and reinforced concrete piles (second-stage piles). The piles are constructed using a semi-interlocking method, where the later-constructed second-phase piles cut into the earlier-constructed, lower-strength first-phase piles, ensuring an interlocking depth of at least 250mm. Simultaneously, the optimization algorithm ensures that the bottom of this curtain is effectively embedded at least 1.5 meters into the underlying relatively impermeable layer (such as clay), forming a reliable vertical water-stop barrier. Subsequently, on the inner side of this curtain, i.e., the excavation face of the foundation pit, for areas requiring soil reinforcement to resist deformation caused by excavation unloading, the MJS method piles are designed. The MJS method pile is an improved, more controllable high-pressure jet grouting pile. Its core process parameters, including the unique horizontal jet nozzle's spray radius, spray pressure up to 40MPa, cement grout spraying duration, and drill rod lifting and rotation speeds, are all calculated and precisely matched by the optimization algorithm based on the physical and mechanical parameters of the silty sand layer to be reinforced (such as the internal friction angle, cohesion, and void ratio provided by the model). The goal is to ensure that the resulting consolidated body is uniform and continuous, and that its unconfined compressive strength stably reaches the design requirements (e.g., not less than 1.5 MPa) after 28 days of curing. These two different pile types are integrated through BIM modeling, spatial layout optimization, and 3D collision checks. This ensures that the water-stopping function of the interlocking piles and the reinforcement function of the MJS piles are spatially coordinated and functionally complementary. This achieves deterministic and high-precision control of groundwater and soil deformation under extremely complex urban geological and environmental conditions.

[0035] Furthermore, to address the technical challenges of traditional cofferdam methods in constructing large bridge pier foundations in turbulent riverbeds or nearshore environments—characterized by high construction difficulty, risk, cost, and significant impact on the aquatic environment—this invention integrates a targeted underwater construction adaptive control module in step four. This module first uses high-precision riverbed topographic data from the three-dimensional geological property model constructed in step one, along with separately input hydrological survey data (such as multi-year average flow velocity and peak flow velocity during flood season), to numerically simulate the complex three-dimensional flow field in the pile foundation construction area using professional fluid dynamics simulation (CFD) software (such as ANSYS Fluent). Based on the simulated velocity and pressure distribution cloud maps, an optimal layout scheme for a temporary diversion system composed of prefabricated heavy steel structure baffles is designed. This scheme aims to reduce the water flow velocity at key pile locations to below the safe operating window (e.g., from the original 2.5 m / s to 0.5 m / s) using the fewest possible components. During construction, the water-retaining plate is first precisely hoisted, placed, and fixed under the guidance of GPS and underwater sonar. After a locally stable water area is formed, a specially designed "three-pipe two-drill method" is used for underwater overburden removal, bedrock drilling, or rock embedding operations. Specifically, the "three-pipe two-drill method" refers to a combined underwater drilling platform consisting of a large-diameter central positioning and casing steel pipe and two symmetrically arranged, slightly smaller-diameter working steel pipes on either side. The central positioning steel pipe is first precisely sunk to the design elevation under the guidance of a three-dimensional positioning system using a high-power vibratory pile hammer, serving as the absolute benchmark for all subsequent operations. Then, the two working steel pipes are lowered into their guide frames, each containing a down-the-hole drill rig driven by a hydraulic power station. Key parameters such as torque, drilling pressure, and drilling speed during the drilling process of these two down-the-hole drill rigs are monitored and analyzed in real time by the construction process dynamic monitoring unit 30 described in step three. When monitoring data shows that the drill bit has encountered hard, slightly weathered bedrock, the adaptive closed-loop control unit 40 instructs the drilling rig to pause drilling. Operators then lower precisely metered and packaged underwater emulsion explosives through the work pipe for controlled micro-yield blasting. This efficiently breaks the rock or creates a rock-embedded anchoring groove on the rock surface that meets design requirements. The entire complex underwater operation is remotely monitored and guided by a 3D visualization platform integrating BIM models, real-time sensor data, and underwater video monitoring, significantly improving construction accuracy, efficiency, and inherent safety under harsh hydrogeological conditions.

[0036] Furthermore, this invention also provides a BIM-based complex geological pile-laying construction system for performing the above-described method. (See reference...) Figure 2The system structure diagram shown illustrates that the system is logically and physically composed of several functional units that are both clearly defined and closely coordinated. These units include: a geological information fusion and modeling unit 10, a pile foundation performance simulation and optimization design unit 20, a construction process dynamic monitoring unit 30, and an adaptive closed-loop control unit 40. These core units achieve high-speed and reliable data interconnection through a low-latency, high-bandwidth communication network based on fiber optics and a 5G industrial private network, ensuring the real-time performance and integrity of the entire information flow from on-site data acquisition, cloud model updates, intelligent decision generation to on-site equipment control.

[0037] As an important complement to the sustainability and environmental friendliness of the entire closed-loop system, the system can also integrate a waste mud resource utilization module 50. A large amount of waste mud is generated during pile foundation construction, especially bored pile construction. This module 50 aims to transform this traditionally considered construction waste into valuable engineering materials. Its specific components include: online monitoring sensors for mud properties, such as online densitometers, rotational viscometers, and pH meters, installed on the mud tank or mud treatment equipment pipelines; an automatic flocculant / solidifier dosing system consisting of a high-precision metering pump, storage tank, and forced mixer; and a portable testing module (such as a micro-penetrator) for rapidly detecting the mechanical properties of solidified soil. The workflow of module 50 is as follows: First, the physicochemical properties (such as density, solids content, viscosity, etc.) of the waste mud discharged from the construction site are monitored in real time. Based on these real-time parameters, a built-in algorithm instantly calculates the optimal dosage of flocculant (such as polyacrylamide PAM) and solidifier (such as 32.5 grade ordinary Portland cement, carbide slag, or fly ash) required to achieve mud-water separation and subsequent solidification. Subsequently, the automated dosing system precisely adds these agents to the mud and mixes them. After a certain period of reaction and curing, the waste mud is transformed into solidified soil with specific load-bearing capacity. This module performs random mechanical property tests on it, for example, ensuring that its 7-day unconfined compressive strength is not less than 0.5 MPa. Finally, the volume, location, and mechanical property parameters of these qualified solidified soils are digitally written back into the project's BIM model as a new, usable "material resource" for unified digital scheduling and management in subsequent site backfilling, roadbed filling, or temporary road paving. The addition of this module not only solves the environmental problem of waste mud treatment and reduces the cost of off-site disposal, but also creates new value through resource utilization, truly realizing a green closed loop for pile foundation engineering.

[0038] Example 1 This example is applied to the pile foundation engineering of a super high-rise office building in Pudong New Area, Shanghai, China. The site has extremely complex geological conditions, mainly composed of Quaternary loose sediments, from top to bottom as follows: 0 to -3.0m is artificial fill; -3.0 to -12.5m is gray silty clay, fluid plastic, with high water content and low bearing capacity; -12.5 to -21.0m is gray silt interbedded with thin layers of silty clay, containing multiple lenses of varying thickness, with locally high confined water head; -21.0 to -45.0m is dark green clay, plastic to stiff plastic, which is the main bearing layer for the pile foundation; below -45.0m is a dense gravel layer. The project design uses bored cast-in-place piles, with a main pile diameter of 1000mm.

[0039] This embodiment employs the method described in this invention for construction. Step 1: First, detailed survey reports of 25 boreholes within the site were collected, and a professional unit was commissioned to conduct high-density resistivity imaging (HRI) exploration covering the entire site, with a survey line spacing of 15 meters and an electrode spacing of 5 meters. The two types of data were input into the geological information fusion and modeling unit 10, and a three-dimensional geological property volume model with voxel dimensions of 1.5m × 1.5m × 0.5m was constructed using a co-kriging interpolation algorithm. This model clearly delineates the three-dimensional spatial morphology and distribution range of silt lenses within a depth range of -12.5 to -21.0m. Step 2: The superstructure BIM model was coupled with this geological model, and analysis was performed in the pile foundation performance simulation and optimization design unit 20. Optimization was performed using a multi-objective genetic algorithm, with decision variables including pile length, pile diameter, and fine-tuning of pile position. Optimization results showed that the pile length of most piles could be optimized from the original uniform 42 meters to 38-41 meters, and in areas with less influence from silty sand lenses, the pile diameter could be reduced from 1000mm to 900mm. The final optimized scheme included 320 piles, of which 280 were 1000mm in diameter and 40 were 900mm in diameter. The pile length was customized according to the burial depth of the bearing layer and the superstructure load. Steps three and four: During construction, each rotary drilling rig was equipped with a monitoring system including sensors for torque, rotational speed, drilling pressure, and depth. When drilling to pile B-17, at a depth of -18.2m, the torque sensor reading suddenly spiked from an average of 120kN·m to 220kN·m, far exceeding the model prediction. The system immediately performed inversion analysis and determined that there was an unexplored, small-scale, high-density cemented gravel mass at this location. The data assimilation module then updated the geological model for this local area. Based on the updated model, the risk prediction module of the adaptive closed-loop control unit 40 determined that the probability of encountering the hard mass at a similar depth with the adjacent pile B-18 under construction exceeded 85%. Therefore, before the construction of pile B-18, the system automatically generated targeted construction instructions, recommending that the operator switch the drilling mode from "standard soil and rock mode" to "hard rock mode" after drilling to -17.5m, increasing torque output reserve in advance and reducing the drilling rate. Simultaneously, when drilling to -20.5m, i.e., when the model predicted that the drilling was about to enter the confined water-bearing silty sand layer, the system automatically instructed the mud pump to increase its output pressure by 10%, and the mud specific gravity from 1.18 g / cm³. 3 Increased to 1.22 g / cm³ 3 .

[0040] Comparative Example 1 This comparative example uses traditional construction methods to construct the same project described in Example 1. Design phase: Geological assessment was based solely on two-dimensional profiles of 25 boreholes, simplifying different soil layers into horizontal homogeneous layers. To ensure safety, a conservative design strategy was adopted, with all 320 piles designed to have a uniform diameter of 1000mm and a uniform length of 42 meters, ensuring all pile tips penetrated sufficiently into the bearing stratum. Construction phase: Layout was conducted using a traditional total station, relying on manual operation. The construction process depended entirely on the experience and judgment of the site engineer and drilling operator. While drilling to pile B-17, at a depth of -18.2m, the drill bit suddenly encountered an unexpected hard lump, causing severe vibration of the drill rod and jamming. It took nearly four hours to resolve the issue and caused some wear to the drill bit. During subsequent drilling of multiple piles to around -20.5m, due to the failure to predict the pressure head of the water in advance and adjust the mud parameters, three minor instances of borehole wall instability and water inrush occurred. Although no major accidents were caused, additional borehole cleaning and more mud were required to stabilize the borehole wall, resulting in project delays and increased costs.

[0041] Comparison of key indicators between Example 1 and Comparative Example 1 The quantitative comparison data in the table above clearly demonstrates that, compared to traditional methods, the BIM-based pile driving construction method described in this invention significantly saves engineering materials and costs during the design phase through refined simulation and optimization. During the construction phase, dynamic model correction and adaptive control effectively mitigate risks, improve construction efficiency, and ensure project quality. Ultimately, it achieves superior settlement control in the structural performance after completion. This invention, through systematic innovation in its technological system, achieves a comprehensive improvement in economic, social, and environmental benefits.

Claims

1. A complex geological pile placing construction method based on BIM technology, characterized in that, The method comprises the following steps: step One: constructing a three-dimensional geological attribute body model, specifically, collecting multi-source heterogeneous geological information including at least existing drilling survey data, wide-area geophysical exploration data and regional geological history data of the site, and using a spatial interpolation algorithm through a geostatistical modeling engine (10) to perform three-dimensional grid subdivision on the target geological space to form a voxel set, and calculating and assigning a set of probabilistic physical and mechanical parameter attributes to each voxel, thereby generating a full-parameterized three-dimensional geological attribute body model capable of representing the attribute of the continuous medium of the underground space; step two: performing integrated performance simulation and collaborative optimization design, specifically, coupling the building information model of the superstructure and the three-dimensional geological attribute body model in a unified coordinate system to form a complete calculation model including the superstructure load, pile foundation structure and geological foundation, using a finite element analysis solver (20) to load and simulate the coupled model, and on this basis, calling an optimization algorithm to iteratively optimize and calculate the decision variables including the length, diameter, type and position of the single pile, until the pile foundation layout and parameter scheme that meets the preset constraint conditions and makes the multi-objective optimization function reach the optimal solution set is found, and a pile foundation BIM model is generated accordingly; step three: implementing dynamic correction and assimilation of the geological attribute body model, specifically, real-time acquisition of construction process parameters through a sensor array (30) arranged on the pile foundation construction equipment, analysis of the construction process parameters into equivalent mechanical parameters of the rock stratum at the position of the drill bit through a geological parameter inversion analysis model, and using a data assimilation module (40) to update and correct the original probabilistic parameters of the corresponding voxels in the three-dimensional geological attribute body model by taking the equivalent mechanical parameters as new observation values; step four: generating and issuing adaptive construction control instructions, specifically, based on the three-dimensional geological attribute body model dynamically corrected in step three, on the one hand, outputting the accurate three-dimensional spatial coordinates of each pile in the pile foundation BIM model into a high-precision lofting data file recognizable by the construction surveying equipment, and on the other hand, using a feedforward risk prediction and control module (40) to simulate the pile segment to be constructed in advance to assess the construction risk, and when the predicted risk probability exceeds the preset safety threshold, automatically generating adaptive construction control instructions for adjusting the construction process parameters, and issuing the instructions to the PLC of the construction equipment for execution.

2. The complex geological pile driving construction method based on BIM technology according to claim 1, characterized in that, In the step one, the multi-source heterogeneous geological information specifically includes: firstly, existing borehole investigation data of the site, which contains the rock-soil layer boundary, rock-soil category, standard penetration test blow count at different depths of each borehole, and water content, void ratio, liquid-plastic limit, compression modulus, internal friction angle and cohesion obtained through indoor soil test; secondly, wide-area geophysical exploration data, which are three-dimensional data field of apparent resistivity or shear wave velocity covering the entire construction area obtained by high-density resistivity method or surface wave exploration method; and the spatial interpolation algorithm is the collaborative Kriging interpolation algorithm, which takes the borehole investigation data as hard data constraint and the geophysical exploration data as soft data trend guidance for interpolation calculation.

3. The complex geological pile driving construction method based on BIM technology according to claim 1, characterized in that, In the step two, the specific implementation of the integrated performance simulation and collaborative optimization design is that the finite element analysis solver (20) generates a finite element grid of the geological part based on the voxel grid of the three-dimensional geological attribute body model, and directly inputs the physical and mechanical parameters carried by each voxel as the material properties of the corresponding finite element unit; the target variables of the multi-objective optimization function include at least the total construction cost, the overall structure safety factor and the maximum differential settlement value; the constraint conditions include various bearing capacity and deformation restrictions required by national specifications; and the optimization algorithm is genetic algorithm or particle swarm optimization algorithm, which takes the single pile length, pile diameter, pile type, pile reinforcement ratio and fine-tuned coordinates of pile position within the allowed range as decision variables for iterative optimization until the Pareto optimal solution set is found.

4. The complex geological pile driving construction method based on BIM technology according to claim 1, characterized in that, In the step three, the sensor array (30) arranged on the pile foundation construction equipment at least includes: a torque-speed sensor for measuring the torque and rotation speed of the drill bit; a displacement-speed sensor for measuring the drilling depth and rate; a pressure-flow sensor for measuring the inlet and outlet pressures and flow rates of the mud circulation system; and an online spectral analyzer or image recognition module based on machine vision for analyzing the rock properties of drill cuttings.

5. The complex geological pile driving construction method based on BIM technology according to claim 4, characterized in that, In the step three, the specific implementation of dynamic correction and assimilation of the geological attribute body model is that the sensor array (30) continuously collects construction process parameters at a preset sampling frequency, and pre-processes and packages them through an edge computing gateway, and uploads them to the central server in real time via an industrial wireless network; the geological parameter inversion analysis model running on the server side inversely analyzes the input construction parameters into the equivalent mechanical parameters of the rock stratum at the position of the drill bit based on the theoretical relationship between rock mechanics and borehole engineering; and the data assimilation module (40) uses Kalman filtering or ensemble Kalman filtering algorithm to take the equivalent mechanical parameters obtained by inversion as new observations to update and correct the original probabilistic parameters of the corresponding voxels in the three-dimensional geological attribute body model.

6. The complex geological pile driving construction method based on BIM technology according to claim 1, characterized in that, In the fourth step, the generated adaptive construction control instructions specifically include at least one of the following: instructions for adjusting the drilling pressure and rotation speed of the rotary drilling rig to adapt to changes in the hardness of the stratum; instructions for adjusting the output flow and pressure of the mud pump to dynamically balance the water and soil pressure of the hole wall; instructions for automatically adjusting the mud proportioning before entering a specific risk stratum.

7. The complex geological pile driving construction method based on BIM technology according to claim 1, characterized in that, When the three-dimensional geological attribute body model reveals that there is a water-rich silt stratum in the site or deep foundation pit excavation needs to be carried out beside an existing building (structure), the step two of the cooperative optimization design further includes: optimizing a composite support and water stop structure scheme formed by MJS method piles and drilled encasement piles; specifically, first, a drilled encasement pile water stop curtain composed of plain concrete piles and reinforced concrete piles is designed, which is half-encased between the piles and embedded in the underlying impermeable layer; then, MJS method piles are used for soil reinforcement on the inside of the curtain, and the optimization algorithm is used to accurately calculate the process parameters of the MJS method piles, such as the horizontal rotary jet radius, jet pressure, jet duration and lifting speed, according to the physical and mechanical parameters of the soil to be reinforced, to ensure that the consolidated body formed reaches the design required unconfined compressive strength.

8. The complex geological pile construction method based on BIM technology according to claim 1, characterized in that, When the construction scene is an underwater environment such as a swift riverbed, the fourth step further includes an underwater construction adaptive control module operation process: first, according to the riverbed topographic data in the three-dimensional geological attribute body model and hydrological survey data, the flow field distribution is calculated through fluid mechanics simulation software, and a diversion system composed of prefabricated steel structure water stop plates is designed and installed according to the calculation results; Then, a "three-pipe two-drill method" is used for underwater operation, which uses a central positioning steel pipe and two symmetrically arranged operation steel pipes to form a combined drilling platform. After the central positioning steel pipe is accurately positioned, a down-the-hole drill is lowered into the operation steel pipe for drilling. During the drilling process, the drilling parameters are monitored in real time. When the monitoring data shows that hard rock is encountered, the adaptive control system instructs the drill to stop and accurately measures the underwater special emulsion explosive to be lowered for micro-explosion to break the rock or form an anchor slot for pile rock-socketing.

9. The complex geological pile driving method based on BIM technology according to claim 1, wherein, When the engineering target is to construct a vertical impermeable barrier, the method further includes: in step two, based on the three-dimensional distribution map of the permeability coefficient in the three-dimensional geological attribute body model, the spatial range, depth and thickness of the impermeable wall are accurately determined, and through the material proportioning optimization algorithm, the grouting material formula is determined, which is based on modified kaolin clay and composite addition of sodium-based bentonite and carboxymethyl cellulose sodium; in the fourth step, the grouting operation is carried out in a constant pressure, constant quantity and constant time manner, and the adaptive control system compares the real-time feedback grouting pressure with the theoretical water and soil pressure in the geological model to dynamically adjust the injection rate, so as to prevent split grouting or insufficient compaction, thereby constructing an impermeable wall with a permeability coefficient reaching the preset low permeability standard.

10. A complex geological pile placing construction system based on BIM technology, characterized in that, Comprise: A geological information fusion and modeling unit (10) internally integrated with standardized data acquisition interfaces, collaborative Kriging geostatistics modeling engine and three-dimensional voxel model database, configured to receive multi-source heterogeneous geological information including site existing borehole survey data, wide-area geophysical exploration data and regional geological history data, and generate a full-parameter three-dimensional geological attribute body model; a pile foundation performance simulation and optimization design unit (20) including high-performance computing server clusters in hardware and solidified with finite element analysis solver coupled with BIM and geological model and multi-objective genetic optimization algorithm program in software, bidirectional data communication with the geological information fusion and modeling unit (10), configured to perform single pile-group pile-cap integrated performance simulation and output optimized pile foundation BIM model; a construction process dynamic monitoring unit (30) including a sensor array laid on the construction equipment for collecting drilling torque, speed, depth, rate and mud pressure, flow and other parameters, an edge computing gateway and a geological parameter inversion analysis module, configured to collect, process and upload construction process parameters in real time and invert them into equivalent mechanical parameters of strata; an adaptive closed-loop control unit (40) internally integrated with Kalman filter data assimilation module, feedforward risk prediction module and PLC control instruction generation module, data interaction with other units, configured to dynamically correct the three-dimensional geological attribute body model according to the equivalent mechanical parameters fed back by the construction process dynamic monitoring unit (30), generate high-precision lofting data and adaptive construction control instructions based on the corrected model, and issue the instructions to the actuators of the construction equipment; a waste mud resource utilization module (50) including mud property online monitoring sensors, automatic flocculating agent dosing system and solidified soil mechanical property detection module, configured to monitor the properties of waste mud in real time, automatically add the optimal dose of flocculating agent and solidifying agent, convert waste mud into solidified soil with specific bearing capacity, and write the parameters back to the BIM model for unified scheduling management.

Images