Rapid lead hole construction process for bare rock geology steel sheet pile cofferdam

Abstract

The invention relates to the technical field of bridge engineering, in particular to a bare rock geology steel sheet pile cofferdam rapid guide hole construction technology which comprises the steps of early-stage geological investigation and design, equipment type selection and arrangement, positioning and paying off, drilling operation, hole cleaning and quality inspection and connection steel sheet pile inserting and driving. Compared with the prior art that hole guiding parameter design mainly depends on artificial experience and scattered geological survey reports, and the risk that decision basis is insufficient, and parameters are conservative or rash exists, the digital design method combining multiple modes of survey with three-dimensional geological modeling and an intelligent prediction model is adopted, and the method has the advantages of being high in practicability and easy to popularize. Through multi-source data fusion, rock mass quality automatic grading and construction process simulation, fundamental transformation of lead hole parameters from empirical judgment to data-driven optimization is realized, so that high adaptability of a design scheme and complex bare rock geological conditions is ensured from the source, the scientificity and the foresight of design are remarkably improved, and the method is suitable for large-scale popularization and application. And a solid foundation is laid for subsequent efficient and accurate construction.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering technology, and in particular to a rapid pre-hole drilling construction process for steel sheet pile cofferdams in bare rock geology. Background Technology

[0002] In water conservancy projects, bridge construction, and port and wharf projects, steel sheet pile cofferdams are widely used as a common temporary water-retaining structure. Their construction quality and efficiency directly affect the safety, progress, and cost of the main project. Especially when constructing cofferdams in bare rock geological conditions, the traditional steel sheet pile driving process faces significant challenges due to the high strength and undulating nature of the rock strata and the lack of a soil cover layer for buffering.

[0003] Currently, the conventional method for constructing sheet pile cofferdams in bare rock geology is to first perform pilot drilling to create a channel for the implantation of sheet piles into the rock strata. However, existing pilot drilling techniques have significant shortcomings. First, the design of pilot drilling parameters relies heavily on the personal experience of engineers, lacking a scientific basis for precise quantification of complex rock conditions. This often leads to overly conservative or aggressive parameter selection, affecting the quality and cost-effectiveness of the borehole. Second, the drilling process is mainly operated manually based on experience, lacking real-time and precise monitoring and correction methods, making it prone to quality defects such as borehole inclination and misalignment due to uneven rock strata or equipment vibration. Third, the connections between various construction procedures are loose, with a long time interval between borehole formation and pile insertion, increasing the risk of borehole wall collapse. Simultaneously, the impermeability of backfill materials and the adaptability of pile driving techniques to bare rock geology are insufficient, affecting the overall water-stopping effect and stability of the cofferdam. Furthermore, construction platforms are mostly fixed structures, making dynamic adjustments difficult to compensate for construction deviations. These shortcomings collectively result in low construction efficiency, high costs, and poor quality control in bare rock geology.

[0004] To address the aforementioned technical challenges, this invention proposes a rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geology. Summary of the Invention

[0005] To overcome the problems mentioned in the background art, the present invention proposes a rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geology.

[0006] The technical solution of this invention is: a rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geology, comprising the following steps: S11: Preliminary geological survey and design: Determine the borehole parameters based on the geological conditions of the bare rock, including borehole diameter, depth and spacing; S12: Equipment selection and layout: Select the pilot hole equipment, install and debug the intelligent construction base on the construction platform, and install the pilot hole equipment after debugging; S13: Positioning and setting out: Use a total station and GPS positioning system to calibrate the center point of the pilot hole; S14: Drilling operation: Start the pilot hole equipment to penetrate the rock surface at a low speed, gradually accelerate to the rated parameters, and monitor the drilling pressure, rotation speed and cuttings in real time. S15: Hole cleaning and quality inspection: Immediately after drilling, clean the hole with high-pressure airflow and clean water, and check the verticality and hole diameter; S16: Connecting steel sheet pile driving: Insert steel sheet piles within 2 hours after the pilot hole is completed, and backfill with anti-seepage material at the same time.

[0007] As a preferred option, the preliminary geological survey and design should include the following steps: S21: Multi-source geological data acquisition: Acquire multi-source heterogeneous geological data of the construction area, including aerial remote sensing data, ground survey data and underwater topographic data; S22: 3D reconstruction of geological model and rock mass quality classification: Import the collected multi-source data into the 3D geological modeling platform for data fusion and 3D geological model reconstruction; and based on the reconstructed model, use the rock mass quality classification standard to automatically partition and classify the rock mass within the model range, generating a rock mass quality partition cloud map. S23: Intelligent borehole parameter decision: The obtained three-dimensional geological model and rock mass quality zoning cloud map are input into the trained intelligent prediction model, and the intelligent prediction model outputs the optimal borehole parameters. S24: Construction Feasibility Simulation and Parameter Optimization: Based on the obtained pilot hole parameters, the construction process is simulated on a digital simulation platform, including drill bit-rock interaction, borehole deviation risk, and rock mass stability between adjacent holes; and the pilot hole parameters are optimized based on the simulation results.

[0008] Preferably, the intelligent prediction model used is a multi-task regression prediction system based on machine learning, and the intelligent prediction model includes: A11: Data Input Interface Module: Used to receive and standardize data from the 3D geological model. Input feature vectors include: rock mass quality classification, rock uniaxial compressive strength, rock mass integrity coefficient, groundwater status, and borehole design depth. A12: Core prediction engine module: It consists of machine learning algorithms trained on historical engineering data, and outputs multiple key aperture parameters in parallel. The machine learning algorithm used is the gradient boosting decision tree algorithm. A13: Results Output and Interpretation Module: Outputs recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated drilling efficiency, and provides confidence intervals and analysis of key influencing factors for each output value.

[0009] Preferably, the intelligent prediction model includes the following steps when it is in operation: S31: Offline Model Training: Construct an ensemble learning model based on gradient boosting decision tree algorithm, and train it offline using no less than 1000 sets of historical engineering datasets. Each set of data contains geological feature vectors and their corresponding successful borehole parameter vectors. Before training, feature engineering is performed on the data, including one-hot encoding of categorical geological features and standardization scaling of numerical features. Logarithmic transformation is performed on the borehole efficiency target variable to optimize the data distribution. S32: Online Prediction and Multi-Objective Optimization: Real-time geological feature vectors from previous explorations, including rock mass quality classification, uniaxial compressive strength of rock, rock mass integrity coefficient, groundwater status, and borehole design depth, are input into the trained model. The model's core prediction engine processes the input data in parallel using a multi-task regression approach. It performs weighted calculations by simultaneously optimizing a multi-objective function that optimizes borehole efficiency, construction cost, and anchorage reliability, and outputs a set of recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated borehole efficiency. S33: Results Interpretation and System Integration: By integrating the SHAP model interpretability component, the influencing factors of the pre-recommended value set are quantitatively analyzed, and a confidence interval and a visualization report of key geological influencing factors for each recommended value are generated and output. At the same time, the final recommended value set is automatically pushed to the construction management BIM platform through the application programming interface, and is directly used to drive the parameter configuration of the automated drilling rig on site. S34: Online Learning and Model Optimization: During the construction of sheet pile driving, real-time data on actual drilling efficiency and hole quality assessment are collected as feedback and compared with the initial prediction values ​​of the model. When the deviation continues to exceed the preset threshold, the online learning mechanism of the model is automatically triggered, and the model parameters are fine-tuned using an incremental learning algorithm to achieve adaptive and continuous optimization of the model's prediction performance in specific projects.

[0010] Preferably, the drilling operation also includes: S41: Real-time data acquisition: Real-time data of the borehole location is acquired through a multi-sensor integrated system, including location coordinates and morphological parameters. The multi-sensor integrated system includes a GPS sensor, a total station, an inclination sensor, and a laser rangefinder. S42: Deviation Analysis: Real-time analysis of the collected data, focusing on the verticality of the hole position, the hole diameter, and the deviation from the calibrated center point of the pilot hole; S43: Correction: Input the deviation data into the intelligent construction base, and perform correction operations through the intelligent construction base.

[0011] Preferably, the deviation analysis process includes: A21: Verticality assessment: Calculate the angle between the actual borehole axis and the designed vertical line; A22: Aperture consistency check: Compare the real-time aperture with the design value; A23: Position offset calculation: Analyze the planar deviation between the borehole center and the calibration point through an algorithm.

[0012] As a preferred option, when drilling, the pilot hole depth is greater than the designed embedment depth of the sheet pile and is 1.5 times the height of the cofferdam. After the hole cleaning operation, the verticality, hole diameter and rock cuttings data of each hole are recorded.

[0013] Preferably, when driving the connecting sheet piles, the backfill material used is a mixture of silty clay, bentonite and sawdust, and the method of inserting the sheet pile is one of a vibratory hammer and a hydraulic impact device, which is slowly pressed in along the guide hole.

[0014] Preferably, the selected drilling equipment includes one of down-the-hole drills, rotary drilling rigs, and impact drills.

[0015] Preferably, the intelligent construction base includes a fixed base, connecting rods, moving rails, and a rail-mounted vehicle. Two sets of fixed bases are respectively installed on construction platforms erected on opposite sides of the sheet pile cofferdam. Multiple sets of connecting rods are configured with a multi-segment structure, and these segments are connected by threads. Both ends of each connecting rod are fixedly connected to the two sets of fixed bases. Multiple sets of moving rails are configured, each with multiple sets of fixing holes. All moving rails are fitted onto the connecting rods through these fixing holes. The rail-mounted vehicle is mounted on the moving rails.

[0016] Preferably, a leveling assembly is provided below the fixed base. The leveling assembly includes an upper fixing ring, a fixed base, a lower fixing ring, and hydraulic rods. Multiple sets of upper fixing rings are provided and are located on the ground of the fixed base. Multiple sets of lower fixing rings are provided and are located on the top surface of the fixed base. Multiple sets of hydraulic rods are provided, and each set of hydraulic rods has a connecting ring at both ends. One end of each set of hydraulic rods is connected to the upper fixing ring through the connecting ring, and the other end of each set of hydraulic rods is connected to the lower fixing ring through the connecting ring. A trackless electric flatbed cart is provided below both sets of fixed bases, and the trackless electric flatbed cart is located below the fixed base.

[0017] As a preferred option, the intelligent construction base also includes an intelligent controller, the working principle of which is as follows: S51: Receive deviation data, read and convert the deviation data, and convert the deviation data into the angle deviation and position deviation of the drilling device; S52: Retrieve status data of multiple sets of hydraulic rods, trackless electric flatcars, and rail-mounted moving vehicles; S53: Calculate the required adjustment amount for multiple sets of hydraulic rods, trackless electric flatcars, and track moving vehicles based on the angular and positional deviations of the drilling equipment; S54: Convert the calculated control quantity into control commands and output them to multiple sets of hydraulic rods, trackless electric flatcars and rail-mounted moving vehicles.

[0018] The beneficial effects of this invention are: 1. Compared with existing technologies that mainly rely on manual experience and scattered geological survey reports for borehole parameter design, which carries the risk of insufficient decision-making basis and conservative or overly aggressive parameters, this invention adopts a digital design method that combines multiple survey methods with three-dimensional geological modeling and intelligent prediction models. Through multi-source data fusion, automatic rock mass quality classification, and construction process simulation, it realizes a fundamental shift from experience-based judgment to data-driven optimization of borehole parameters. This ensures the high adaptability of the design scheme to complex bare rock geological conditions from the source, significantly improves the scientific nature and foresight of the design, and lays a solid foundation for subsequent efficient and precise construction. 2. Compared with the existing technology where the drilling process mainly relies on the operator's experience for control and lacks real-time monitoring and active correction methods, which easily leads to quality problems such as hole tilt and deviation, this invention introduces a real-time monitoring system integrating multiple sensors and an intelligent construction base, forming a closed-loop control process of measurement, analysis and adjustment. It can dynamically sense the drilling posture and automatically and accurately correct deviations, thereby upgrading the control of hole quality from passive acceptance to active guarantee, greatly improving the vertical accuracy and positional accuracy of drilling, effectively avoiding rework and ensuring the success rate of hole formation on the first attempt. 3. Compared with existing technologies that typically use standard depths or empirical formulas to determine the embedment depth, which may affect the stability of the cofferdam due to insufficient anchoring force in bare rock geology, this invention explicitly specifies that the borehole depth is 1.5 times the cofferdam height. This ultra-deep design significantly increases the embedment length of the sheet piles in the rock strata. Combined with the system recording the borehole quality data after cleaning, a closed-loop quality traceability system is formed, which greatly enhances the overturning safety and overall stability of the cofferdam structure, providing a key guarantee for safe and reliable bottomless construction under harsh hydrological conditions. 4. Compared to existing construction platforms, which are typically fixed structures, making it difficult to adapt to complex sites and unable to dynamically adjust equipment position to compensate for construction deviations, this invention designs an intelligent construction platform consisting of a fixed base, connecting rods, moving rails, leveling components, moving devices, and an intelligent controller. This achieves the stability and adjustability of the construction platform in terms of mechanical structure. Furthermore, through intelligent algorithms, the real-time monitored deviations are transformed into precise multi-degree-of-freedom control commands for the platform, thereby constructing a dynamic operating system that can actively maintain drilling accuracy. This fundamentally improves the adaptability and precision of the process in complex environments. Attached Figure Description

[0019] Figure 1 The diagram shows a flow chart of the rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geological conditions according to the present invention. Figure 2 The diagram shows the process flow of preliminary geological survey and design in the rapid pilot hole construction technology of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 3 The diagram shows a three-dimensional structural schematic of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 4 The diagram shown is a side perspective view of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 5 The diagram shown is a side perspective view of the fixed base of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 6 The diagram shown is a three-dimensional structural schematic of the other side of the fixed base of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 7 The diagram shows a three-dimensional side view of the moving track of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 8 The diagram shows another side view of the moving track of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 9 The diagram shows a three-dimensional side view of the connecting rod of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Figure 10 The diagram shows the working principle of the intelligent controller of the intelligent construction base in the rapid pilot hole construction process of steel sheet pile cofferdam in bare rock geology of the present invention. Explanation of reference numerals in the attached drawings: 1. Fixed base; 2. Connecting rod; 3. Moving track; 4. Track moving vehicle; 5. Fixing hole; 6. Leveling assembly; 601. Upper fixing ring; 602. Fixed base; 603. Lower fixing ring; 604. Hydraulic rod; 605. Connecting ring; 7. Trackless electric flatcar. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] Please see Figure 1 - Figure 2 This invention provides an embodiment: a rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geology, comprising the following steps: S11: Preliminary geological survey and design: Determine the borehole parameters based on the geological conditions of the bare rock, including borehole diameter, depth and spacing; S12: Equipment selection and layout: Select the pilot hole equipment, install and debug the intelligent construction base on the construction platform, and install the pilot hole equipment after debugging; S13: Positioning and setting out: Use a total station and GPS positioning system to calibrate the center point of the pilot hole; S14: Drilling operation: Start the pilot hole equipment to penetrate the rock surface at a low speed, gradually accelerate to the rated parameters, and monitor the drilling pressure, rotation speed and cuttings in real time. S15: Hole cleaning and quality inspection: Immediately after drilling, clean the hole with high-pressure airflow and clean water, and check the verticality and hole diameter; S16: Connecting steel sheet pile driving: Insert steel sheet piles within 2 hours after the pilot hole is completed, and backfill with anti-seepage material at the same time.

[0022] In this embodiment, the present invention proposes a rapid pilot-hole construction process for steel sheet pile cofferdams in bare rock geology. The core of this process involves the following steps: First, preliminary geological surveys and design are conducted to determine key pilot-hole parameters; then, equipment selection and layout are performed, and an intelligent construction base is installed and debugged on the construction platform to ensure precise equipment positioning; next, a total station and GPS are used for high-precision positioning and layout; during drilling, the speed is gradually increased to the rated parameters after penetrating the rock surface at a low speed, with real-time monitoring of the drilling status throughout; immediately after hole formation, high-pressure airflow and clean water are used for hole cleaning, and verticality and hole diameter are strictly inspected; finally, steel sheet piles are driven and anti-seepage materials are backfilled within 2 hours after pilot-hole completion. This process, through streamlined and intelligent construction organization, significantly improves construction efficiency and hole formation quality under complex geological conditions of bare rock. The application of the intelligent construction base ensures the accuracy and stability of equipment installation; the strict hole cleaning and quality inspection effectively avoids hole collapse and deviation; and the tight process connection (pile driving within 2 hours) greatly reduces the risk of hole wall instability, thus achieving rapid, accurate, and reliable cofferdam construction overall.

[0023] As a preferred option, the preliminary geological survey and design should include the following steps: S21: Multi-source geological data acquisition: Acquire multi-source heterogeneous geological data of the construction area, including aerial remote sensing data, ground survey data and underwater topographic data; S22: 3D reconstruction of geological model and rock mass quality classification: Import the collected multi-source data into the 3D geological modeling platform for data fusion and 3D geological model reconstruction; and based on the reconstructed model, use the rock mass quality classification standard to automatically partition and classify the rock mass within the model range, generating a rock mass quality partition cloud map. S23: Intelligent borehole parameter decision: The obtained three-dimensional geological model and rock mass quality zoning cloud map are input into the trained intelligent prediction model, and the intelligent prediction model outputs the optimal borehole parameters. S24: Construction Feasibility Simulation and Parameter Optimization: Based on the obtained pilot hole parameters, the construction process is simulated on a digital simulation platform, including drill bit-rock interaction, borehole deviation risk, and rock mass stability between adjacent holes; and the pilot hole parameters are optimized based on the simulation results.

[0024] In this embodiment, the present invention deepens and innovates the preliminary geological survey and design stage. Its core lies in executing a closed-loop digital design chain: first, multi-source geological data is collected through integrated air, ground, and water surveys; then, a 3D geological modeling platform is used to fuse the data and reconstruct the geological model, achieving automated zoning and grading of rock mass quality; subsequently, the geological model and zoning results are input into a trained intelligent prediction model, outputting optimal borehole parameters in a data-driven manner; finally, the construction process is simulated through a digital simulation platform, with parameter feedback optimization and risk prediction. This scheme elevates traditional surveying to an intelligent decision-making and simulation verification process based on a precise geological model. Its beneficial effects are a fundamental improvement in the scientific rigor and foresight of the design. It not only ensures a high degree of compatibility between borehole parameters and complex bare rock geological conditions through intelligent algorithms, achieving a leap from "experience-based design" to "data-driven optimization," but also avoids potential risks such as borehole deviation and rock mass instability through construction simulation, thereby guaranteeing the feasibility, efficiency, and economy of subsequent construction from the source, laying a solid foundation for the rapid and precise construction of the entire cofferdam project.

[0025] Preferably, the intelligent prediction model used is a multi-task regression prediction system based on machine learning, and the intelligent prediction model includes: A11: Data Input Interface Module: Used to receive and standardize data from the 3D geological model. Input feature vectors include: rock mass quality classification, rock uniaxial compressive strength, rock mass integrity coefficient, groundwater status, and borehole design depth. A12: Core prediction engine module: It consists of machine learning algorithms trained on historical engineering data, and outputs multiple key aperture parameters in parallel. The machine learning algorithm used is the gradient boosting decision tree algorithm. A13: Results Output and Interpretation Module: Outputs recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated drilling efficiency, and provides confidence intervals and analysis of key influencing factors for each output value.

[0026] In this embodiment, the intelligent prediction model employed by the present invention is a multi-task regression prediction system based on machine learning. Its core technology lies in constructing an intelligent decision-making architecture that includes a data input interface, a core prediction engine, and a result output and interpretation module. The model receives and standardizes key geological feature vectors such as rock mass quality grading and uniaxial compressive strength through the data input interface module. Subsequently, the core prediction engine, built based on a gradient boosting decision tree algorithm, performs parallel computation, outputting recommended values ​​for multiple key borehole parameters, such as target borehole diameter, borehole depth, borehole spacing, and drilling efficiency, all at once. Finally, the result output and interpretation module not only provides parameter suggestions but also simultaneously gives the confidence interval for each parameter and a visual analysis of the main influencing factors. This technology represents a leap from traditional experience-based judgment to data-driven intelligent computing in borehole parameter decision-making. By using machine learning algorithms to accurately uncover the nonlinear relationship between complex geological conditions and construction parameters, it significantly improves the scientific rigor and accuracy of parameter setting. Its multi-task parallel output mode ensures optimal synergy among various parameters, while its unique interpretability analysis function empowers engineers to understand the basis of their decisions, significantly enhancing the transparency and credibility of decisions. This fundamentally guarantees the efficiency and success rate of subsequent borehole construction.

[0027] Preferably, the intelligent prediction model includes the following steps when it is in operation: S31: Offline Model Training: Construct an ensemble learning model based on gradient boosting decision tree algorithm, and train it offline using no less than 1000 sets of historical engineering datasets. Each set of data contains geological feature vectors and their corresponding successful borehole parameter vectors. Before training, feature engineering is performed on the data, including one-hot encoding of categorical geological features and standardization scaling of numerical features. Logarithmic transformation is performed on the borehole efficiency target variable to optimize the data distribution. S32: Online Prediction and Multi-Objective Optimization: Real-time geological feature vectors from previous explorations, including rock mass quality classification, uniaxial compressive strength of rock, rock mass integrity coefficient, groundwater status, and borehole design depth, are input into the trained model. The model's core prediction engine processes the input data in parallel using a multi-task regression approach. It performs weighted calculations by simultaneously optimizing a multi-objective function that optimizes borehole efficiency, construction cost, and anchorage reliability, and outputs a set of recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated borehole efficiency. S33: Results Interpretation and System Integration: By integrating the SHAP model interpretability component, the influencing factors of the pre-recommended value set are quantitatively analyzed, and a confidence interval and a visualization report of key geological influencing factors for each recommended value are generated and output. At the same time, the final recommended value set is automatically pushed to the construction management BIM platform through the application programming interface, and is directly used to drive the parameter configuration of the automated drilling rig on site. S34: Online Learning and Model Optimization: During the construction of sheet pile driving, real-time data on actual drilling efficiency and hole quality assessment are collected as feedback and compared with the initial prediction values ​​of the model. When the deviation continues to exceed the preset threshold, the online learning mechanism of the model is automatically triggered, and the model parameters are fine-tuned using an incremental learning algorithm to achieve adaptive and continuous optimization of the model's prediction performance in specific projects.

[0028] In this embodiment, the intelligent prediction model of the present invention is first trained offline based on a large amount of historical data (no less than 1,000 sets), and the data quality is ensured through meticulous feature engineering. During actual prediction, it can simultaneously optimize multiple objectives such as drilling efficiency, cost, and anchoring reliability, and output the optimal set of pilot hole parameters in one go. Then, it uses the SHAP model to provide decision interpretation and confidence intervals, and automatically integrates the results into the BIM platform to drive on-site equipment. Finally, it triggers an online learning mechanism through real-time construction feedback data to achieve adaptive and continuous optimization of the model in specific projects. This constructs an "intelligent brain" with both forward-looking decision-making and dynamic self-optimization capabilities. It not only ensures the scientific and economical nature of pilot hole parameters from the source through data-driven and multi-party balanced optimization, greatly improving the construction success rate and reducing costs, but also enhances the transparency and credibility of engineering decisions through the interpretability of results. Most importantly, its unique online learning capability enables the model to continuously improve itself according to actual working conditions, thereby significantly improving the adaptability and long-term reliability of the entire process to complex bare rock geological conditions, and realizing a fundamental leap from static experience-based to dynamic intelligence in construction decision-making.

[0029] Preferably, the drilling operation also includes: S41: Real-time data acquisition: Real-time data of the borehole location is acquired through a multi-sensor integrated system, including location coordinates and morphological parameters. The multi-sensor integrated system includes a GPS sensor, a total station, an inclination sensor, and a laser rangefinder. S42: Deviation Analysis: Real-time analysis of the collected data, focusing on the verticality of the hole position, the hole diameter, and the deviation from the calibrated center point of the pilot hole; S43: Correction: Input the deviation data into the intelligent construction base, and perform correction operations through the intelligent construction base.

[0030] In this embodiment, the present invention introduces a real-time monitoring and active correction system in the drilling operation. The core technology lies in: using a multi-sensor system composed of GPS sensors, total stations, tilt sensors, and laser rangefinders to collect real-time, high-precision data on the spatial coordinates and morphological parameters of the borehole position during the drilling process; then analyzing the collected data, calculating and displaying the verticality, diameter, and deviation from the designed borehole position in real time; finally, feeding back the deviation data to the intelligent construction base in real time, which automatically or assisted the operator in performing the correction operation. The beneficial effect of this solution is that it achieves a qualitative leap in drilling operations from "result acceptance" to "process controllability." Through real-time data-driven closed-loop control of "measurement-analysis-adjustment," it can dynamically suppress deviations and errors caused by factors such as uneven rock strata and equipment vibration. This transforms the control of hole quality from passive to active, thereby greatly ensuring the vertical and positional accuracy of the drilling. This lays a solid foundation for the smooth and accurate driving of subsequent sheet piles, effectively avoids rework caused by excessive hole deviation, and significantly improves the first-time success rate and overall efficiency of construction.

[0031] Preferably, the deviation analysis process includes: A21: Verticality assessment: Calculate the angle between the actual borehole axis and the designed vertical line; A22: Aperture consistency check: Compare the real-time aperture with the design value; A23: Position offset calculation: Analyze the planar deviation between the borehole center and the calibration point through an algorithm.

[0032] In this embodiment, the deviation analysis process specifically includes three precise quantitative evaluation steps: verticality assessment by calculating the angle between the actual borehole axis and the designed vertical line; borehole diameter consistency check by comparing the real-time borehole diameter with the design value; and position offset calculation by analyzing the planar deviation between the borehole center and the calibration point using an algorithm. The beneficial effect of this analysis process is that it transforms traditional qualitative judgment into quantitative, multi-dimensional diagnosis based on real-time data. By simultaneously and accurately measuring the three key quality indicators—verticality, borehole diameter, and planar position—a comprehensive borehole quality evaluation system is constructed. This allows for the rapid and objective identification of the specific type and degree of construction deviation, providing precise data support and a clear adjustment direction for subsequent intelligent correction. This fundamentally ensures that the borehole quality meets the stringent design requirements and effectively avoids potential engineering quality hazards caused by the loss of control over a single indicator.

[0033] As a preferred option, when drilling, the pilot hole depth is greater than the designed embedment depth of the sheet pile and is 1.5 times the height of the cofferdam. After the hole cleaning operation, the verticality, hole diameter and rock cuttings data of each hole are recorded.

[0034] In this embodiment, the present invention specifies two key quality control measures during drilling operations: first, the pilot hole depth is set to be greater than the designed embedment depth of the sheet piles, specifically quantified as 1.5 times the cofferdam height; second, after hole cleaning, the verticality, diameter, and cuttings data of each hole are systematically recorded. The beneficial effects of this approach are that by setting ultra-deep pilot holes of 1.5 times the cofferdam height, the embedment length of the sheet piles in the bare rock is significantly increased, thereby greatly improving the overturning stability and overall safety of the cofferdam structure, providing a fundamental guarantee for achieving "bottomless construction" under complex hydrogeological conditions. Simultaneously, the complete recording of borehole quality data not only achieves traceability of the construction process but, more importantly, provides valuable quantitative data support for subsequent construction parameter optimization and similar projects, constructing a closed-loop quality management system of "construction-recording-optimization," thereby continuously promoting the precision and intelligent development of construction technology. Preferably, when driving the connecting sheet piles, the backfill material used is a mixture of silty clay, bentonite and sawdust, and the method of inserting the sheet pile is one of a vibratory hammer and a hydraulic impact device, which is slowly pressed in along the guide hole.

[0035] In this embodiment, the invention uses a special mixture composed of silty clay, bentonite, and sawdust as backfill material, and employs a vibratory hammer or hydraulic impact equipment to slowly drive the sheet piles into the pre-drilled borehole. This achieves the following: the special backfill mixture fully utilizes the synergistic effect of its components; the silty clay forms the main body, the bentonite expands upon contact with water to effectively block seepage, and the sawdust fills the gaps to improve workability, collectively ensuring excellent anti-seepage performance under short seepage paths in bare rock; simultaneously, the selection of suitable pile driving equipment and adherence to the principle of "slowly driving in along the guide" greatly reduces disturbance to the pre-formed borehole wall, effectively overcomes rock strata obstacles, ensures the sheet piles are accurately driven to the design elevation, and ultimately jointly guarantees the reliability of the cofferdam structure's water-stopping mechanism and overall stability.

[0036] Preferably, the selected drilling equipment includes one of down-the-hole drills, rotary drilling rigs, and impact drills.

[0037] In this embodiment, three specialized equipment options are provided for borehole equipment selection: down-the-hole drills (DHD), rotary drilling rigs, and impact drills, suitable for different rock strata conditions. DHD, with its high-frequency impact characteristics, efficiently breaks hard rocks such as moderately weathered sandstone; rotary drilling rigs are suitable for applications requiring high borehole wall stability and core sampling, achieving efficient and clean borehole formation; while impact drills demonstrate powerful penetration capabilities when dealing with extremely hard or complex rock strata with large fractures. This targeted equipment selection strategy ensures that the construction team can choose the most suitable equipment based on the specific bare rock geological conditions, thereby fundamentally guaranteeing the efficiency, borehole quality, and economy of the drilling operation, laying a solid foundation for the smooth implementation of the entire rapid construction process.

[0038] Please see Figure 3 - Figure 10 The intelligent construction base includes a fixed base 1, connecting rods 2, moving rails 3, and a rail-moving vehicle 4. Two sets of fixed bases 1 are respectively installed on construction platforms built on opposite sides of the steel sheet pile cofferdam. Multiple sets of connecting rods 2 are configured with a multi-segment structure, and the segments of the connecting rods 2 are connected by threads. Both ends of the connecting rods 2 are fixedly connected to the two sets of fixed bases 1. Multiple sets of moving rails 3 are configured, each with multiple sets of fixing holes 5. All sets of moving rails 3 are fitted onto the connecting rods 2 through the fixing holes 5. The rail-moving vehicle 4 is mounted on the moving rails 3.

[0039] Preferably, a leveling assembly 6 is provided below the fixed base 1. The leveling assembly 6 includes an upper fixing ring 601, a fixed base 602, a lower fixing ring 603, and hydraulic rods 604. Multiple sets of upper fixing rings 601 are provided and are located on the ground of the fixed base 1. Multiple sets of lower fixing rings 603 are provided and are located on the top surface of the fixed base 602. Multiple sets of hydraulic rods 604 are provided, and each end of the multiple sets of hydraulic rods 604 is provided with a connecting ring 605. One end of the multiple sets of hydraulic rods 604 is connected to the upper fixing ring 601 through the connecting ring 605, and the other end of the multiple sets of hydraulic rods 604 is connected to the lower fixing ring 603 through the connecting ring 605. A trackless electric flatbed cart 7 is provided below both sets of fixed bases 1, and the trackless electric flatbed cart 7 is located below the fixed base 602.

[0040] As a preferred option, the intelligent construction base also includes an intelligent controller, the working principle of which is as follows: S51: Receive deviation data, read and convert the deviation data, and convert the deviation data into the angle deviation and position deviation of the drilling device; S52: Retrieve status data of multiple sets of hydraulic rods 604, trackless electric flatcar 7 and track moving car 4; S53: Calculate the required adjustment amount for multiple sets of hydraulic rods 604, trackless electric flatcar 7 and track moving car 4 based on the angle deviation and position deviation of the drilling equipment; S54: Convert the calculated control quantity into control commands and output them to multiple sets of hydraulic rods 604, trackless electric flatcars 7 and track moving cars 4.

[0041] The intelligent construction base is composed of a mechanical structure and an intelligent control system. Its mechanical components include fixed bases 1 on both sides of the construction platform, connected by multi-segment threaded connecting rods 2, a movable track 3 with fixing holes 5 fitted onto the connecting rods 2, and a track-mounted mobile vehicle 4 carrying the construction equipment. Below the fixed bases 1 is a leveling assembly 6 consisting of a hydraulic rod 604 connecting an upper fixing ring 601 and a lower fixing ring 603, and the entire assembly is supported by a trackless electric flatbed vehicle 7. The core intelligent controller's workflow is as follows: it receives deviation data and converts it into the angle and position deviation of the equipment; it retrieves the status data of the hydraulic rod 604, the trackless electric flatbed vehicle 7, and the track-mounted mobile vehicle 4; it calculates the required precise adjustment amount; and it outputs control commands to drive the actuators. This has led to the construction of a construction platform that integrates mechanical stability, multi-degree-of-freedom adjustment capabilities, and intelligent feedback control. It can automatically and precisely adjust the position and attitude of the drilling rig based on real-time monitoring of drilling deviations, thereby proactively suppressing construction deviations at the source. This upgrades the traditional static and passive construction platform into a dynamic and proactive precision operation system, greatly ensuring the verticality and positional accuracy of borehole operations under complex bare rock geological conditions, and providing a solid hardware foundation for the entire rapid construction process.

[0042] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology, characterized in that: Includes the following steps: S11: Preliminary geological survey and design: Determine the borehole parameters based on the geological conditions of the bare rock, including borehole diameter, depth and spacing; S12: Equipment selection and layout: Select the pilot hole equipment, install and debug the intelligent construction base on the construction platform, and install the pilot hole equipment after debugging; S13: Positioning and setting out: Use a total station and GPS positioning system to calibrate the center point of the pilot hole; S14: Drilling operation: Start the pilot hole equipment to penetrate the rock surface at a low speed, gradually accelerate to the rated parameters, and monitor the drilling pressure, rotation speed and cuttings in real time. S15: Hole cleaning and quality inspection: Immediately after drilling, clean the hole with high-pressure airflow and clean water, and check the verticality and hole diameter; S16: Connecting steel sheet pile driving: Insert steel sheet piles within 2 hours after the pilot hole is completed, and backfill with anti-seepage material at the same time.

2. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geological conditions according to claim 1, characterized in that: The preliminary geological survey and design process includes the following steps: S21: Multi-source geological data acquisition: Acquire multi-source heterogeneous geological data of the construction area, including aerial remote sensing data, ground survey data and underwater topographic data; S22: 3D reconstruction of geological model and rock mass quality classification: Import the collected multi-source data into the 3D geological modeling platform for data fusion and 3D geological model reconstruction; and based on the reconstructed model, use the rock mass quality classification standard to automatically partition and classify the rock mass within the model range, generating a rock mass quality partition cloud map. S23: Intelligent borehole parameter decision: The obtained three-dimensional geological model and rock mass quality zoning cloud map are input into the trained intelligent prediction model, and the intelligent prediction model outputs the optimal borehole parameters. S24: Construction Feasibility Simulation and Parameter Optimization: Based on the obtained pilot hole parameters, the construction process is simulated on a digital simulation platform, including drill bit-rock interaction, borehole deviation risk, and rock mass stability between adjacent holes; and the pilot hole parameters are optimized based on the simulation results.

3. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 2, characterized in that: The intelligent prediction model used is a multi-task regression prediction system based on machine learning, and the intelligent prediction model includes: A11: Data Input Interface Module: Used to receive and standardize data from the 3D geological model. Input feature vectors include: rock mass quality classification, rock uniaxial compressive strength, rock mass integrity coefficient, groundwater status, and borehole design depth. A12: Core prediction engine module: It consists of machine learning algorithms trained on historical engineering data, and outputs multiple key aperture parameters in parallel. The machine learning algorithm used is the gradient boosting decision tree algorithm. A13: Results Output and Interpretation Module: Outputs recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated drilling efficiency, and provides confidence intervals and analysis of key influencing factors for each output value.

4. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 3, characterized in that: When an intelligent prediction model is working, it includes the following steps: S31: Offline Model Training: Construct an ensemble learning model based on gradient boosting decision tree algorithm, and train it offline using no less than 1000 sets of historical engineering datasets. Each set of data contains geological feature vectors and their corresponding successful borehole parameter vectors. Before training, feature engineering is performed on the data, including one-hot encoding of categorical geological features and standardization scaling of numerical features. Logarithmic transformation is performed on the borehole efficiency target variable to optimize the data distribution. S32: Online Prediction and Multi-Objective Optimization: Real-time geological feature vectors from previous explorations, including rock mass quality classification, uniaxial compressive strength of rock, rock mass integrity coefficient, groundwater status, and borehole design depth, are input into the trained model. The model's core prediction engine processes the input data in parallel using a multi-task regression approach. It performs weighted calculations by simultaneously optimizing a multi-objective function that optimizes borehole efficiency, construction cost, and anchorage reliability, and outputs a set of recommended values ​​for target borehole diameter, suggested borehole depth, borehole spacing, and estimated borehole efficiency. S33: Results Interpretation and System Integration: By integrating the SHAP model interpretability component, the influencing factors of the pre-recommended value set are quantitatively analyzed, and a confidence interval and a visualization report of key geological influencing factors for each recommended value are generated and output. At the same time, the final recommended value set is automatically pushed to the construction management BIM platform through the application programming interface, and is directly used to drive the parameter configuration of the automated drilling rig on site. S34: Online Learning and Model Optimization: During the construction of sheet pile driving, real-time data on actual drilling efficiency and hole quality assessment are collected as feedback and compared with the initial prediction values ​​of the model. When the deviation continues to exceed the preset threshold, the online learning mechanism of the model is automatically triggered, and the model parameters are fine-tuned using an incremental learning algorithm to achieve adaptive and continuous optimization of the model's prediction performance in specific projects.

5. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geological conditions according to claim 4, characterized in that: Drilling operations also include: S41: Real-time data acquisition: Real-time data of the borehole location is acquired through a multi-sensor integrated system, including location coordinates and morphological parameters. The multi-sensor integrated system includes a GPS sensor, a total station, an inclination sensor, and a laser rangefinder. S42: Deviation Analysis: Real-time analysis of the collected data, focusing on the verticality of the hole position, the hole diameter, and the deviation from the calibrated center point of the pilot hole; S43: Correction: Input the deviation data into the intelligent construction base, and perform the correction operation through the intelligent construction base.

6. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 5, characterized in that: The analysis process of deviation analysis includes: A21: Verticality assessment: Calculate the angle between the actual borehole axis and the designed vertical line; A22: Aperture consistency check: Compare the real-time aperture with the design value; A23: Position offset calculation: Analyze the planar deviation between the borehole center and the calibration point through an algorithm.

7. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 6, characterized in that: During drilling operations, the pilot hole depth is greater than the designed embedment depth of the sheet pile and is 1.5 times the height of the cofferdam. After the hole cleaning operation, the verticality, diameter and rock cuttings data of each hole are recorded.

8. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 7, characterized in that: When driving sheet piles for connection, the backfill material used is a mixture of silty clay, bentonite and sawdust, and the method of inserting the sheet pile is one of vibratory hammer and hydraulic impact equipment, which is slowly pressed in along the pilot hole guide.

9. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 1, characterized in that: The intelligent construction base includes a fixed base (1), connecting rods (2), moving rails (3) and a rail moving vehicle (4). The fixed bases (1) are set in two groups, and the two groups of fixed bases (1) are respectively set on the construction platforms built on opposite sides of the steel sheet pile cofferdam. The connecting rods (2) are set in multiple groups, and the connecting rods (2) are set in a multi-segment structure. The multi-segment structure of the connecting rods (2) is connected by threads. The two ends of the connecting rods (2) are respectively fixedly connected to the two groups of fixed bases (1). The moving rails (3) are set in multiple groups, and multiple sets of fixing holes (5) are opened on the multiple groups of moving rails (3). The multiple groups of moving rails (3) are all sleeved on the connecting rods (2) through the fixing holes (5). The rail moving vehicle (4) is set on the moving rails (3).

10. The rapid pilot-hole construction technology for steel sheet pile cofferdams in bare rock geology according to claim 9, characterized in that: A leveling assembly (6) is provided below the fixed base (1). The leveling assembly (6) includes an upper fixing ring (601), a fixed base (602), a lower fixing ring (603), and hydraulic rods (604). Multiple sets of upper fixing rings (601) are provided, and the upper fixing rings (601) are located on the ground of the fixed base (1). Multiple sets of lower fixing rings (603) are provided, and the lower fixing rings (603) are located on the top surface of the fixed base (602). Multiple sets of hydraulic rods (604) are provided. Both ends of the hydraulic rod (604) are provided with connecting rings (605). One end of the multiple sets of hydraulic rods (604) is connected to the upper fixed ring (601) through the connecting ring (605), and the other end of the multiple sets of hydraulic rods (604) is connected to the lower fixed ring (603) through the connecting ring (605). A trackless electric flatbed cart (7) is provided below both sets of fixed bases (1), and the trackless electric flatbed cart (7) is located below the fixed base (602).

Images