High-precision leveling construction method for deep-water foundation groove rubble pier

Abstract

The application discloses a high-precision leveling construction method for a deep-water foundation groove gravel pier, and belongs to the technical field of marine engineering construction. The method solves the problems of low efficiency, poor precision, high risk, and the incapability of real-time evaluation of the filling body state and the difficulty in coping with the dynamic underwater environment caused by the operation of divers in the deep-water foundation groove. The technical scheme comprises the following steps: accurately positioning and parking a construction ship; lowering a leveling frame provided with inertial navigation and a sonar height measuring instrument and remotely adjusting the leveling by hydraulic pressure; lowering a telescopic gravel pouring funnel integrated with a camera and a flow rate sensor to pour gravel, and automatically adjusting the length of a guide pipe and the gravel pouring position based on the real-time collected elevation, flatness, flow rate and tidal level data; automatically controlling a high-frequency vibration scraping mechanism to scrape the gravel; and finally, scanning and accepting. The method is mainly used for the high-precision leveling construction of a large-area gravel foundation (such as a immersed tunnel foundation and a gravity wharf bed) in a deep-water environment.

Description

Technical Field

[0001] This invention relates to the field of marine engineering construction technology. More specifically, this invention relates to a high-precision leveling construction method for deep-water foundation trenches and crushed stone piers, which is particularly suitable for high-precision leveling construction of large-area crushed stone foundations such as immersed tunnel foundations and gravity wharf foundations. Background Technology

[0002] In the construction of deep-water foundation trenches and crushed stone piers, traditional leveling methods have long relied on underwater operations by divers. Divers must descend into deep water to assess and measure the surface condition of the crushed stone fill using visual observation and underwater measuring tools. This method has several significant limitations. First, manual underwater measurement is inefficient, and the results are significantly affected by the diver's individual experience, underwater visibility, and physical condition, leading to strong subjectivity, poor repeatability, and difficulty in ensuring accuracy. Second, surface commanders cannot directly perceive the real-time status of the underwater stone placement and can only make decisions based on verbal descriptions or intermittent communication from divers. This experience-based remote operation lacks objective data support, making it difficult to optimize and adjust construction parameters in real time.

[0003] Traditional methods lack real-time, quantitative assessment tools for the morphology of underwater fill embankments. During construction, it's impossible to continuously acquire elevation and flatness data of the fill surface; only after the completion of individual piers can a workboat equipped with a multibeam echo sounder perform post-construction scanning and inspection. If the inspection reveals substandard flatness, remedial measures are required, significantly increasing construction costs and causing delays. This post-construction inspection model makes the construction process lack controllability, hindering true process quality control.

[0004] Another prominent issue is that traditional construction methods struggle to effectively address the dynamic effects of deep-water environments. The speed, direction, and constant changes in underwater currents and tide levels significantly impact the trajectory and accumulation pattern of crushed stone. However, traditionally, the stone placement rate is often set as a fixed value based on experience, failing to dynamically adjust to variations in water flow. This leads to the crushed stone easily dispersing under strong currents or piling up excessively in weak currents, resulting in poor uniformity and uneven compaction of the fill material, directly affecting the overall stability and long-term durability of the crushed stone pier.

[0005] Therefore, there has long been an urgent need in this field for a leveling construction method that can overcome the above-mentioned limitations and is applicable to large-area deep-water trench crushed stone piers, so as to achieve high-precision construction control of the underwater fill morphology. Summary of the Invention

[0006] One object of the present invention is to solve at least the above-mentioned problems and to provide at least the advantages that will be described later.

[0007] To achieve these objectives and other advantages according to the present invention, a high-precision leveling construction method for deep-water foundation trench crushed stone piers is provided, comprising:

[0008] S1. The construction vessel moves to the design axis position of the construction pier based on the RTK-GPS positioning signal and anchors in place.

[0009] S2. The leveling frame, equipped with an inertial navigation unit and a sonar altimeter, is lowered above the construction pier of the seabed trench. During the lowering, the distance between the frame and the bottom of the trench is monitored in real time using the sonar altimeter. After being lowered to the foundation bed, the multiple legs of the leveling frame are remotely and hydraulically controlled through the control console based on the three-dimensional attitude data fed back by the inertial navigation unit and the height data fed back by the sonar altimeter, so that the top surface reaches the design elevation and remains horizontal.

[0010] S3. Deploy a retractable rock-throwing funnel. Its guide tube integrates an underwater camera and a flow velocity sensor. Adjust the distance between the funnel outlet and the bed surface to 1.0-2.0m, and throw rocks into the leveling frame through the funnel. During rock throwing, the underwater camera captures and processes images of the rock spread and accumulation, obtaining data on the elevation and flatness of the throwing surface. The flow velocity sensor monitors the flow velocity at the funnel outlet. The tide level sensor monitors the tide level. The integrated control system receives data on the elevation, flatness, flow velocity, and tide level of the throwing surface. Based on the tide level and the real-time elevation of the throwing surface, it automatically adjusts the length of the guide tube to maintain the vertical distance between the funnel outlet and the throwing surface at 1.0-2.0m. Based on the flatness data, it controls the movement of the ship's anchoring and positioning system to adjust the rock-throwing point and replenish rock materials to low-lying areas.

[0011] S4. After the stone material covers the top surface of the leveling frame, the high-frequency vibration scraping mechanism integrated into the leveling frame is activated to scrape off the stone material exceeding the top surface. The start, stop and operation of the high-frequency vibration scraping mechanism are automatically controlled by the integrated control system. The steps include: the integrated control system receives real-time flatness data processed by the underwater camera. When the average elevation of the filling surface reaches the design elevation of the top surface of the leveling frame and the rate of change of the real-time flatness data is lower than the preset threshold, the system automatically issues a start command. The high-frequency vibration scraping mechanism scrapes back and forth along the preset track at a fixed speed. When the integrated control system detects that the flatness deviation of the top surface of the pier does not exceed the preset difference threshold, the operation is automatically stopped.

[0012] Preferably, it also includes:

[0013] S5. After completing the slab throwing and leveling of a single rubble block, lift the leveling frame and the retractable slab throwing funnel.

[0014] S6. Use a workboat equipped with a multibeam echo sounder to scan, inspect and accept the constructed crushed stone piers;

[0015] S7. Repeat steps S1-S6 to construct the next construction pier until all construction piers are completed.

[0016] Preferably, before lowering the leveling frame, a multibeam echo sounder mounted on a workboat is used to conduct a full-coverage scan of the seabed trench to verify the elevation of the excavation surface of the seabed trench.

[0017] Preferably, in the stone-throwing step, the crushed stone used has a particle size of 10-50mm and a continuous gradation.

[0018] Preferably, in step S3, the integrated control system dynamically adjusts the stone conveying rate based on the real-time flow velocity data at the funnel outlet monitored by the flow velocity sensor. Specifically, the integrated control system has a pre-stored correspondence table between flow velocity and conveying rate, which specifies the stone conveying rate value corresponding to different flow velocity ranges. The integrated control system queries the correspondence table and automatically adjusts the stone conveying rate to the corresponding target value according to the range of real-time flow velocity.

[0019] Preferably, the integrated control system also includes a data recording and remote transmission module, which is used to record sensor data, control commands and equipment status information throughout the construction process, and transmit the sensor data, control commands and equipment status information to the shore-based monitoring center in real time via a satellite communication link.

[0020] Preferably, in step S2, pressure sensors are installed at the bottom of multiple legs of the leveling frame. After the top surface of the leveling frame is adjusted to the design elevation and kept horizontal, the integrated control system continuously monitors the pressure data at the bottom of each leg. If the pressure value of a single leg decreases by more than a first preset threshold, and the pressure value of one or more other legs increases by more than a second preset threshold, and the horizontal tilt angle data of the leveling frame fed back by the inertial navigation unit exceeds a third preset threshold, the integrated control system determines that the foundation under that leg has become unstable or eroded. The integrated control system issues an alarm and automatically suspends subsequent rock-throwing operations.

[0021] Preferably, in step S3, the specific steps for acquiring the elevation and flatness data of the dumping surface in real time include continuously acquiring underwater video streams during the dumping process using an underwater high-definition camera integrated on the retractable dumping funnel guide tube at a preset frame rate threshold, and transmitting the streams in real time to the integrated control system via armored optical cable. The integrated control system then performs the following processing on the received image sequences sequentially:

[0022] Each frame of the image undergoes processing including histogram equalization and physical model-based scattering effect elimination to enhance image contrast and edge features.

[0023] Based on binocular vision or structured light 3D reconstruction algorithms, matching the same feature points in the left and right images, and generating a 3D point cloud dataset representing the current morphology of the dumping surface through triangulation.

[0024] The 3D point cloud data is registered with the design coordinate system of the leveling frame, and a continuous digital elevation model is generated through gridding and interpolation algorithms.

[0025] Based on the digital elevation model, the average elevation of the entire filling area and the flatness data used to characterize the degree of surface undulation are calculated in real time.

[0026] The present invention has at least the following beneficial effects:

[0027] First, this invention effectively overcomes the problems of traditional deep-water leveling operations relying on manual labor and lacking precision. By integrating multi-source sensors and automatic control technology, it achieves data-driven management of the entire process of rock placement and leveling. The precise positioning of the construction vessel and the intelligent leveling of the leveling frame provide a stable benchmark for subsequent operations; the dynamic adjustment of rock placement parameters and positions based on real-time sensing data significantly improves the uniformity of rock placement; and the automatic start-stop control of the leveling mechanism ensures the achievement of high standards of flatness on the pier top. This method transforms the invisible underwater construction into a visible, controllable, and optimizable digital process, greatly reducing manual intervention, lowering operational risks, and providing reliable foundation construction quality assurance for deep-water infrastructure.

[0028] Secondly, this invention avoids the risk of leveling frame failure due to uneven base. By conducting a comprehensive and precise measurement of the seabed trench before lowering the leveling frame, accurate and complete base elevation data is obtained, providing a reliable basis for the precise calibration of the leveling frame. This measure effectively prevents secondary adjustments or even rework caused by base discrepancies with the design, improves the initial adjustment success rate, and ensures the benchmark stability of subsequent rock-laying and leveling operations, laying a solid foundation for the entire high-precision construction from the initial stage.

[0029] Third, this invention ensures the internal structural quality of the crushed stone pier from a material perspective. By strictly controlling the particle size range of the crushed stone and ensuring gradation continuity, particles of different sizes can interlock and effectively fill gaps during the dumping and water flow impact process, thus forming a strong, dense, and structurally uniform dumping body. This optimized material property provides a stable working surface for subsequent leveling operations, significantly improving the overall stability and load-bearing capacity of the formed pier, and avoiding internal weakness and uneven settlement caused by material separation.

[0030] Fourth, by continuously monitoring flatness data and its rate of change, the integrated control system of this invention can accurately determine the optimal timing for starting and stopping the leveling process, avoiding accuracy losses caused by operating too early or too late. The leveling mechanism runs along a preset track at a constant speed, ensuring consistency and full coverage of the operation. The fully automated control process completely replaces the high-risk and inefficient underwater manual operation, eliminates human interference, significantly improves leveling accuracy, efficiency, and operational safety, and guarantees that the high standard of flatness of the pier top is met on the first attempt.

[0031] Fifth, by establishing a matching relationship between flow velocity and stone placement rate, this invention enables the system to sense changes in water flow in real time and automatically adjust the stone delivery rate. This effectively suppresses the dispersion of crushed stone caused by strong water flow and avoids excessive accumulation under weak water flow. This dynamic adjustment strategy ensures that the stone can be accurately deposited in the target area even in changing environments, significantly improving the uniformity and density of the fill material, and ensuring the final construction quality of the crushed stone pier from a process control perspective.

[0032] Sixth, this invention integrates scattered construction data into a structured information flow with time-series tags, and enables remote real-time monitoring from shore via satellite links, completely breaking down information barriers in offshore construction. Management personnel can gain a comprehensive understanding of the site situation, promptly identify and handle anomalies, and the complete electronic logs provide a valuable data foundation for subsequent quality analysis, responsibility identification, and process optimization, greatly improving the modernization level and decision-making efficiency of offshore engineering management.

[0033] Seventh, through the fusion analysis of pressure sensor and attitude data, this invention enables the system to keenly identify signs of instability or erosion in the foundation beneath the outriggers, and automatically suspend operations and issue alarms before an accident occurs. This proactive safety mechanism based on multi-parameter logical judgment transforms traditional passive post-accident remediation into proactive pre-accident prevention, effectively avoiding equipment overturning and engineering quality accidents, and providing reliable technical support for safe and efficient construction under complex seabed geological conditions.

[0034] Eighth, this invention, through image processing and computer vision algorithms, transforms underwater video streams into high-precision digital elevation models, enabling real-time, objective, and quantitative assessment of the elevation and flatness of the fill surface. This system completely replaces subjective judgment relying on divers' experience, providing precise data input for the automatic decision-making of the integrated control system, ensuring the timeliness and scientific nature of construction parameter adjustments, and serving as the technological cornerstone for achieving high-precision intelligent control throughout the entire process.

[0035] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0036] Figure 1 This is a schematic diagram illustrating the workflow of one of the technical solutions of the present invention. Detailed Implementation

[0037] The present invention will now be described in further detail with reference to examples, so that those skilled in the art can implement it based on the description.

[0038] It should be noted that, unless otherwise specified, the experimental methods described in the following embodiments are conventional methods, and the reagents and materials mentioned are commercially available. It should be particularly noted that the method provided by this invention is mainly aimed at crushed stone foundations with large construction areas and high leveling accuracy requirements, such as the foundation cushion layer of immersed tunnels and the riprap foundation bed of gravity wharves. For smaller areas (usually less than 10m²), [the method is not applicable]. 2 The construction of crushed stone blocks (such as some submarine pipeline support points) varies greatly in terms of construction technology, precision requirements and economic efficiency. The construction equipment and process can be simplified according to the actual situation, and will not be described in detail in this invention.

[0039] like Figure 1 As shown, this invention provides a high-precision leveling construction method for deep-water foundation trench crushed stone piers. Addressing the technical challenges of leveling the top surface of foundation trench crushed stone piers in deep-water environments, which is highly difficult and significantly affected by the marine environment, this invention proposes a systematic solution. Traditional construction methods mainly rely on underwater visual observation and manual measurement by divers, which is not only inefficient and difficult to guarantee accuracy, but also poses high risks to divers. Surface commanders cannot monitor the underwater stone placement status in real time and can only rely on experience, unable to dynamically adjust construction parameters according to actual conditions. Furthermore, traditional methods cannot quantitatively assess the elevation and flatness of the underwater fill in real time; inspection can only be carried out after construction is completed, making it difficult to cope with the influence of dynamic environmental factors such as water flow and tide level, resulting in difficulties in controlling the accuracy of stone placement.

[0040] The implementation of this invention first involves using an RTK-GPS positioning system mounted on a construction vessel to precisely locate the axis of the designed pier position and anchor it in place. Then, a leveling frame equipped with an inertial navigation unit and a sonar altimeter is lowered to the predetermined position in the seabed trench. During the lowering process, the sonar altimeter continuously monitors the distance between the leveling frame and the bottom of the trench to ensure safe lowering. After the leveling frame is in place, based on the three-dimensional attitude data fed back by the inertial navigation unit and the height data from the sonar altimeter, multiple outriggers of the leveling frame are remotely and hydraulically adjusted via a deck control console to precisely adjust its top surface to the designed elevation and maintain it level.

[0041] Next, a retractable rock-throwing funnel, integrated with an underwater camera and flow velocity sensor, is lowered, and the distance between its outlet and the bed surface is adjusted to within the range of 1.0-2.0m. During the rock-throwing process, the underwater camera captures real-time images of the rock dispersion and accumulation, and the image processing system extracts the elevation and flatness data of the filling surface in real time. The flow velocity sensor simultaneously monitors the flow velocity changes at the funnel outlet. The integrated control system comprehensively receives the elevation, flatness, and flow velocity data, and automatically adjusts the length of the rock-throwing guide tube based on real-time tide changes and the elevation of the filling surface, dynamically maintaining the distance between the outlet and the rock surface within the set range. Simultaneously, the system controls the ship's anchoring and positioning system to fine-tune the rock-throwing points based on flatness data, replenishing rocks to identified low-lying areas, achieving real-time feedback and optimization of the filling process.

[0042] After the stones cover the top surface of the leveling frame, the high-frequency vibration scraping mechanism integrated on the leveling frame is activated to scrape off any excess stones. After the construction of a single pier is completed, the leveling frame and the stone-throwing funnel are lifted, and finally, a workboat equipped with a multibeam echo sounder is used to conduct a comprehensive scan, inspection, and acceptance of the constructed crushed stone piers.

[0043] Compared to traditional methods, this invention significantly improves the accuracy and efficiency of underwater rock throwing and leveling through multi-sensor integration, real-time data feedback, and automatic control, reduces reliance on divers, enhances operational safety, and effectively addresses the challenges posed by the dynamic deep-water environment.

[0044] An underwater high-definition camera integrated into the retractable rock-throwing funnel guide continuously captures underwater video streams of rock accumulation and diffusion during the rock-throwing process at a certain frame rate. This video data is transmitted in real-time via armored fiber optic cable to an integrated control system on the deck of the construction vessel, where it undergoes the following processing steps:

[0045] 1. Image preprocessing: Each received image frame is preprocessed, including eliminating color distortion and scattering effects caused by water, enhancing contrast, and reducing noise to obtain clearer outlines of gravel particles.

[0046] 2. Feature Point Matching and 3D Point Cloud Generation: 3D reconstruction algorithms such as binocular vision or structured light are employed. If a binocular camera is used, corresponding feature points in the images captured by the two cameras are matched, and the positions of these points in 3D space are calculated using triangulation. Finally, a 3D point cloud dataset representing the current morphology of the filling surface is generated.

[0047] 3. Digital Elevation Model Establishment: The generated 3D point cloud data is registered and aligned with the design coordinate system of the leveling frame. Then, the discrete point cloud data is meshed and interpolated to generate a continuous digital elevation model. This digital elevation model accurately reflects the elevation of each point on the filling surface relative to the leveling frame reference surface.

[0048] 4. Data Extraction and Analysis: Based on the digital elevation model, two key indicators are calculated and output in real time: one is the average elevation of the entire filling area; the other is the flatness data used to measure the degree of surface undulation, such as the maximum elevation difference or root mean square error.

[0049] Based on tidal changes and the real-time elevation of the filling surface, the guide length of the telescopic rock-filling funnel is automatically adjusted. Through real-time sensing, data fusion, and logical judgment, the hydraulic actuator is driven to operate. The control steps include:

[0050] 1. Data Acquisition and Fusion: Continuously receive data from two independent sources: one is real-time tide data sent by the tide sensor installed on the side of the ship; the other is the average elevation data of the dumping area calculated by the above image processing.

[0051] 2. Calculate the target conduit length: The built-in control logic calculates the required absolute water depth position at the conduit outlet using the following formula: Target outlet water depth = Real-time tide level - (Average elevation of the dumping area + Preset hovering height), where the preset hovering height is a fixed value set by engineers, ranging from 1.0-2.0m, designed to optimize the trajectory of the crushed stone and reduce dispersion. Then, based on the target outlet water depth and the fixed installation position of the dumping funnel on the ship, the required target length of the conduit is calculated.

[0052] 3. Length Adjustment Execution: The calculated target length is compared with the current measured length of the conduit. If a deviation exists, a command is sent to the hydraulic station to drive the hydraulic cylinder that adjusts the telescopic conduit, extending or shortening the conduit until its actual length matches the target length. This process is automatically and cyclically performed during all-weather construction, based on tidal fluctuations and the rise of the backfill surface, ensuring that the conduit outlet is always dynamically maintained at the ideal height.

[0053] Based on real-time flatness data, the anchoring and positioning system of the construction vessel is moved to adjust the rock-dropping point. The operation steps include:

[0054] 1. Smoothness Analysis and Low-lying Area Identification: The integrated control system analyzes the real-time generated digital elevation model and visually displays the high and low areas on the filling surface through color rendering or contour line recognition. It also automatically identifies areas whose elevation is lower than the design elevation and whose deviation exceeds the allowable threshold (e.g., 3cm) and marks them as "low-lying areas" to be filled.

[0055] 2. Generate rock-drop point compensation instructions: For each identified low-lying area, calculate its two-dimensional plane coordinates in the leveling frame coordinate system. Then convert these coordinates into offset and direction relative to the construction vessel.

[0056] 3. Vessel Position Fine-tuning: The integrated control system sends the compensation command to the construction vessel's anchoring and positioning system (such as a DGPS dynamic positioning system or an azimuth thruster system). The anchoring and positioning system controls the construction vessel's mooring lines or thrusters, enabling the vessel to make small, precise movements while anchored, aligning the outlet of the retractable riprap funnel with the center coordinates of the identified low-lying area.

[0057] 4. Targeted replenishment: After the construction vessel is moved into position, the stone conveying system is controlled to replenish stones at specific points and in specific quantities in the designated low-lying area. After replenishment is completed, the image processing system scans the area again to verify the replenishment effect. If the requirements are still not met, steps 2-4 are repeated until the flatness of the area meets the standard.

[0058] By cyclically running the above three processes, high precision and high efficiency in leveling crushed stone blocks in deep water environments can be ensured.

[0059] In another embodiment of the present invention, a high-precision leveling construction method for crushed stone piers in deep water foundation trenches is provided. The crushed stone used in the construction is screened and proportioned, with the particle size strictly controlled between 10-50mm, ensuring continuous gradation. Specifically, in the selected crushed stone material, particles with a diameter of 10-20mm account for approximately 30%, particles with a diameter of 20-40mm account for approximately 50%, and particles with a diameter of 40-50mm account for approximately 20%. This continuously graded crushed stone mixture is filled using a retractable stone-throwing funnel. During the impact and descent of the water flow, crushed stone particles of different sizes interlock, effectively reducing the gaps between particles, thereby forming a crushed stone pier with better integrity and a more uniform internal structure. Subsequent leveling operations act on a more integral foundation bed, significantly improving the density and stability of the final crushed stone pier top surface.

[0060] Traditional construction often uses single-size or graded crushed stone, typically with a particle size range of 20-40mm, placed in the same deep water environment and under the same current conditions. Due to the lack of fine particles to fill the gaps between coarse particles and the insufficient proportion of coarse particles to form a skeleton, the fill material itself has high porosity. During the filling process, especially under the scouring effect of water flow, the crushed stone particles lack interlocking force, easily leading to relative displacement and separation. The resulting crushed stone mound has a loose internal structure, with significant differences in density across different areas, resulting in poor overall stability.

[0061] In another embodiment of the present invention, in the traditional deep-water foundation trench crushed stone pier leveling construction, the leveling operation relies entirely on underwater manual operation by divers. Divers must descend underwater and judge the elevation and flatness of the crushed stone pier's top surface by visual observation and experience, manually operating the leveling equipment to perform the leveling operation. This method is greatly affected by underwater visibility, the diver's physical strength and experience, making it difficult to guarantee leveling accuracy, resulting in low efficiency. Furthermore, divers face high safety risks due to prolonged operation in a deep-water, high-pressure environment. Surface commanders cannot monitor the underwater leveling progress and effects in real time, and can only make decisions based on intermittent feedback from divers, leading to poor overall controllability of the construction process.

[0062] In this embodiment, the leveling operation is automatically controlled by an integrated control system, requiring no underwater intervention from divers. The start-up, operation, and shutdown of the high-frequency vibration leveling mechanism integrated on the leveling frame are all automatically managed by the integrated control system. The integrated control system continuously receives real-time flatness data processed by an underwater camera. When the average elevation of the filling surface reaches the design elevation of the top surface of the leveling frame, and the rate of change of the real-time flatness data is less than 1 cm / s (preset rate of change threshold) for 30 consecutive seconds, the system automatically sends a start command to the high-frequency vibration leveling mechanism. The high-frequency vibration leveling mechanism then performs reciprocating leveling operations along the preset track of the leveling frame at a fixed speed. The integrated control system continuously monitors the flatness data during the leveling process. When the flatness deviation of the top surface of the pier is no greater than 30 mm (preset difference threshold), the system automatically stops the leveling operation.

[0063] Compared to traditional methods that rely on manual operation by divers, this invention automates the entire leveling process through an integrated control system, eliminating the uncertainties of human intervention and significantly improving leveling accuracy and consistency. Simultaneously, it avoids underwater operations by divers, greatly reducing construction risks and enhancing overall construction efficiency and safety.

[0064] In another embodiment of the present invention, in the traditional deep-water trench crushed stone pier construction, the stone conveying rate is usually set to a fixed value based on experience. During construction, operators cannot perceive changes in the water flow velocity at the outlet of the retractable crushing hopper in real time. When the ocean current velocity increases, the crushed stone thrown at a fixed rate will be strongly eroded by the water flow, causing some crushed stone, especially fine particles, to be carried away from the target area, resulting in stone loss and uneven distribution. Conversely, when the water flow velocity decreases, the stone at a fixed rate may accumulate excessively below the outlet, forming local bulges. This problem of unevenness of the crushed stone body caused by changes in water flow is difficult to avoid in traditional methods, directly affecting the density and overall stability of the crushed stone pier.

[0065] This implementation achieves dynamic adjustment of the stone conveying rate through an integrated control system. During construction, a flow velocity sensor integrated into the telescopic stone-throwing funnel guide continuously monitors the real-time flow velocity data at the funnel outlet and transmits this data to the integrated control system in real time. The integrated control system has a pre-stored table of correspondence between flow velocity and conveying rate, which defines the optimal stone conveying rate value corresponding to different flow velocity ranges.

[0066] After receiving real-time flow velocity data, the integrated control system immediately queries a built-in correspondence table to determine the current flow velocity range and automatically adjusts the speed of the stone conveying system to the target value corresponding to that range. For example, when an increase in water flow velocity is detected, the system will correspondingly reduce the conveying speed to minimize the impact of water erosion on the crushed stone during its descent; when the water flow velocity decreases, the system will increase the conveying speed to ensure efficient stone placement and prevent excessive accumulation.

[0067] The entire process is fully automated, requiring no human intervention, and can respond instantly to dynamic changes in the water flow environment. Compared to traditional construction methods that use a fixed stone-throwing rate, this invention effectively suppresses the dispersion of crushed stone under the action of water flow by sensing changes in water flow in real time and intelligently adjusting the stone-throwing parameters. This ensures that the stone can be deposited more accurately in the target area, thereby significantly improving the uniformity of the filling and the overall construction quality.

[0068] In another embodiment of this invention, significant deficiencies exist in the recording and management of construction data during traditional deep-water trench and crushed stone pier construction. Various key data, such as altitude data from sonar altimeters, attitude data from inertial navigation units, image data from underwater cameras, flow velocity data from current sensors, and tide level data from tide sensors, are typically recorded separately by their respective independent systems. This data is often stored on local equipment on the construction vessel, managed in a decentralized, offline manner. Operators need to manually record construction logs and subsequently organize and align data from different sources. This model leads to data recording being prone to omissions, asynchrony, and inconsistent formats. More importantly, all this valuable construction process information is isolated at the offshore construction site, inaccessible to shore-based project management and technical teams in a timely manner. They are completely unaware of the real-time location of the construction vessel, the leveling status of the leveling frame, the progress of rock-dropping operations, and the current sea conditions, relying solely on intermittent radio reports from vessel personnel for general understanding. Once quality problems arise, tracing the cause becomes extremely difficult due to the lack of continuous, complete, and timestamped data records to reconstruct the actual construction scenario. Remote monitoring and real-time decision-making guidance are out of the question. The entire construction process is almost a "black box" for modern project management, with huge management blind spots and quality risks.

[0069] This implementation utilizes a built-in data recording and remote transmission module within the integrated control system. This module, a core component of the integrated control system, automatically and continuously collects and records data from all sensors throughout the construction process, all control commands issued by the integrated control system, and the status information of key equipment. All this information is assigned a unified timestamp and packaged into structured data files. Subsequently, through a wide-coverage satellite communication link, this packaged data is transmitted in real-time and continuously to a remote shore-based monitoring center.

[0070] At the shore-based monitoring center, this data is received and analyzed by specialized software. Managers can view the precise location of construction vessels, the 3D posture of the leveling frame, the real-time digital elevation model of the filling surface, the flow velocity at the funnel outlet, the current tide level, and all key parameters such as the stone conveying rate on a large display screen. The entire construction scenario is presented transparently in a visualized and digital manner. Any abnormal data or operations can be detected immediately. All data transmitted to the shore is simultaneously and permanently stored, forming a complete and readily searchable electronic construction log. Every step and every parameter of any crushed stone pier, from start to acceptance, is fully recorded, providing a data foundation for subsequent quality analysis, efficiency optimization, and accountability.

[0071] Compared to traditional methods, this implementation upgrades construction data management from a decentralized, offline, and reactive model to a centralized, online, real-time, and transparent digital management model. It completely breaks down the information barriers between offshore construction and onshore management, enabling shore-based personnel to monitor the construction status as if they were on-site, achieving true remote visual monitoring and full-process quality traceability, and greatly improving the modernization level and controllability of large-scale offshore engineering construction.

[0072] In another embodiment of the invention, the leveling frame is supported and leveled by its legs. If there is a localized weak layer below the seabed trench excavation surface, or if the stability of the trench edge decreases due to excavation disturbance, the legs of the leveling frame may experience uneven settlement during leveling or subsequent rock-filling and leveling operations. This will directly compromise the levelness and design elevation of the leveled frame, causing the entire high-precision construction process to fail, and may even lead to a safety accident caused by the leveling frame overturning. Even if the "elevation" of the trench excavation surface is verified by multibeam echo sounding, this only provides topographic information and cannot reveal the "mechanical properties" (bearing capacity, density) of the soil below the seabed. The construction vessel is unaware of this, making it a risky operation of "blindly" lowering the leveling frame.

[0073] In traditional deep-water trench and gravel pier construction, the stability of the leveling frame after it is seated relies entirely on prior measurements and experience-based judgment of the seabed topography. Construction personnel cannot know in real time the actual bearing capacity of the seabed foundation beneath each leg of the leveling frame. After the leveling frame is leveled, operators are completely unaware of whether the legs have subsided or the foundation has been eroded during the entire rock-dropping and leveling process. If a weak zone exists in the soil beneath a leg, or if cavities are formed due to water erosion, that leg will suddenly sink, causing the entire leveling frame to tilt. This tilt not only immediately destroys the already adjusted design elevation and levelness, rendering all previous high-precision rock-dropping and leveling work meaningless, but more dangerously, it may cause uneven stress on the leveling frame structure, leading to overall overturning and a serious equipment safety accident. Because all operations are carried out in deep water, post-accident remediation and equipment salvage will be extremely difficult and costly. Traditional methods lack effective early warning and immediate response mechanisms for such risks.

[0074] This implementation constructs a real-time foundation stability monitoring system using pressure sensors installed at the bottom of each outrigger of the leveling frame and the inertial navigation unit integrated into the leveling frame. After the leveling frame is leveled, the integrated control system continuously reads the readings from each pressure sensor and the horizontal tilt data from the inertial navigation unit. The integrated control system has several preset thresholds; for example, the first preset threshold corresponds to an abnormal decrease in outrigger pressure, the second preset threshold corresponds to an abnormal increase in pressure on other outriggers, and the third preset threshold corresponds to the maximum permissible horizontal tilt angle of the leveling frame. During construction, if the integrated control system detects that the pressure value of a certain outrigger decreases by more than the first preset threshold within a short period, while the pressure values ​​of one or more adjacent outriggers increase by more than the second preset threshold, and the horizontal tilt data fed back by the inertial navigation unit also exceeds the third preset threshold, the system will immediately perform a logical judgment. This combination of pressure redistribution and attitude change clearly indicates that the foundation beneath that outrigger is becoming unstable.

[0075] Once foundation instability is detected, the integrated control system will immediately and automatically suspend the ongoing rock-dropping or leveling operations and issue an emergency alarm to the operators at the ship's deck control console via audible and visual alarms. Operators can quickly identify the location and type of risk based on the alarm information. Compared to traditional methods that only detect accidents passively after they occur, this implementation method allows the construction team to promptly interrupt operations at the initial stage of an accident, thereby avoiding equipment damage and engineering quality accidents, and gaining valuable time for subsequent remedial measures.

[0076] In another embodiment of this invention, in the construction of traditional deep-water foundation trenches and crushed stone piers, the understanding of the underwater fill morphology relies heavily on manual observation and post-construction inspection by divers. Divers must dive underwater, relying on visual observation and experience to judge the crushed stone accumulation, and obtain limited elevation data through manual measuring tools. This method is greatly affected by underwater visibility, diver physical strength, and subjective experience, making it impossible to conduct a comprehensive, objective, and continuous quantitative assessment of the fill surface morphology. Surface commanders can only make decisions based on intermittent, qualitative descriptions from divers, unable to obtain accurate elevation and flatness data in real time. After construction, a separate workboat equipped with a multibeam echo sounder is required for scanning and acceptance. If unqualified areas are found, the remedial costs are extremely high and the construction period is severely delayed. Throughout the process, the adjustment of construction parameters is severely lagging, unable to cope with the dynamic effects of water flow and tidal changes during the filling process, resulting in poor precision in controlling the flatness of the final crushed stone pier top surface, making it difficult to meet high-standard installation requirements.

[0077] In this embodiment, the specific method for obtaining real-time elevation and flatness data of the fill surface is as follows:

[0078] After the construction vessel completes its positioning and stationing, a retractable rock-throwing funnel, integrated with an underwater high-definition camera, is lowered to the pier construction area. This underwater high-definition camera, with a preset frame rate threshold (10-30 frames per second) and a fixed acquisition frequency, continuously acquires underwater video images of the falling, spreading, and accumulating rocks during the rock-throwing process. The video image data is uploaded in real-time to the integrated control system on the deck of the construction vessel via armored communication fiber optic cable.

[0079] The integrated control system has a built-in image processing unit, which performs the following operations sequentially on each received image frame:

[0080] First, image preprocessing is performed. Preprocessing includes histogram equalization, which expands the contrast range of the image by redistributing the intensity values ​​of pixels, making the details in both light and dark areas more distinct. Preprocessing also includes scattering effect elimination based on a physical optics model. This process digitally corrects color distortion and blurring in the image by establishing a model of light attenuation and scattering in water, restoring the image's true colors and clarity. This enhances the contours and textures of the pebbles in the water, facilitating subsequent machine recognition and matching.

[0081] The preprocessed images are fed into the 3D reconstruction module. This module employs a binocular vision 3D reconstruction algorithm. The algorithm relies on pre-calibrated binocular cameras, whose calibration parameters include internal camera parameters such as focal length, principal point coordinates, and distortion coefficients, as well as external parameters such as the relative position and pose between the two cameras. The algorithm first extracts feature points from the left and right images, using feature description algorithms such as SIFT or ORB to identify key points in the images. Then, feature point matching is performed, pairing feature points extracted from the first camera image with those extracted from the second camera image to find corresponding feature points representing the same physical point. For each successfully matched pair of feature points, its coordinates in 3D space are calculated using triangulation. The triangulation calculation process is based on the perspective projection models of the two cameras and their relative geometric relationship, calculating the true 3D position of the feature point by solving for the pixel coordinate difference (i.e., parallax) between the two images. All calculated 3D points are aggregated to form a dense 3D point cloud dataset representing the surface morphology of the current filling surface.

[0082] Then, a digital elevation model (DEM) is established. The integrated control system registers and aligns the generated 3D point cloud data with the inherent design coordinate system of the leveling frame. The registration process involves finding the correspondence between the point cloud data and known control points in the design coordinate system, and calculating an optimal spatial transformation matrix (usually including rotation and translation). After alignment, the system meshes the discrete 3D point cloud data, dividing the filling area into regular grid cells. For each grid cell, if it contains multiple 3D points, its average elevation is taken; if there are no data points in the grid cell (i.e., gaps exist), an interpolation algorithm, such as inverse distance weighted interpolation or Kriging interpolation, is used to calculate the elevation value of the grid cell based on the elevation values ​​of the surrounding grid points. Through this process, a continuous DEM covering the entire filling area is generated. This DEM is a two-dimensional matrix, where each element represents the elevation at that grid point.

[0083] Finally, data extraction is performed. Based on the generated digital elevation model, the integrated control system calculates the overall average elevation of the filling area in real time. This value is the arithmetic mean of the elevations of all grid points in the model. Simultaneously, the system calculates smoothness data to characterize the degree of surface undulation. For example, it calculates the deviation between the elevations of all grid points and the average elevation, and obtains the root mean square value of these deviations as a smoothness evaluation index; or it directly identifies the highest and lowest points in the model and calculates their difference as the maximum elevation difference.

[0084] Compared to the aforementioned traditional methods, this implementation method completely replaces the outdated approach that relies on divers' visual observation and manual measurement through an automated computer vision and image processing system. It transforms the underwater, invisible filling process into a digital elevation model that can be observed in real-time and precisely quantified on a large screen in the shore-based monitoring center, enabling process control and real-time optimization of construction quality. Based on real-time feedback of elevation and flatness data, managers can instruct the integrated control system to automatically adjust the placement of riprap points or construction parameters, ensuring that the top surface of each final crushed stone pier meets the high-precision flatness standard required by the design on the first attempt, avoiding the huge costs and time losses associated with subsequent remedial measures.

[0085] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. Other modifications can be easily made by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and examples shown and described herein.

Claims

1. A high-precision leveling construction method for crushed stone piers in deep water foundation trenches, characterized in that, include: S1. The construction vessel moves to the design axis position of the construction pier based on the RTK-GPS positioning signal and anchors in place. S2. The leveling frame, equipped with an inertial navigation unit and a sonar altimeter, is lowered above the construction pier of the seabed trench. During the lowering, the distance between the frame and the bottom of the trench is monitored in real time using the sonar altimeter. After being lowered to the foundation bed, the multiple legs of the leveling frame are remotely and hydraulically controlled through the control console based on the three-dimensional attitude data fed back by the inertial navigation unit and the height data fed back by the sonar altimeter, so that the top surface reaches the design elevation and remains horizontal. S3. Deploy a retractable rock-throwing funnel. Its guide tube integrates an underwater camera and a flow velocity sensor. Adjust the distance between the funnel outlet and the bed surface to 1.0-2.0m, and throw rocks into the leveling frame through the funnel. During rock throwing, the underwater camera captures and processes images of the rock spread and accumulation, obtaining data on the elevation and flatness of the throwing surface. The flow velocity sensor monitors the flow velocity at the funnel outlet. The tide level sensor monitors the tide level. The integrated control system receives data on the elevation, flatness, flow velocity, and tide level of the throwing surface. Based on the tide level and the real-time elevation of the throwing surface, it automatically adjusts the length of the guide tube to maintain the vertical distance between the funnel outlet and the throwing surface at 1.0-2.0m. Based on the flatness data, it controls the movement of the ship's anchoring and positioning system to adjust the rock-throwing point and replenish rock materials to low-lying areas. S4. After the stone material covers the top surface of the leveling frame, start the high-frequency vibration scraping mechanism integrated into the leveling frame to scrape off the stone material that exceeds the top surface. The start-up, shutdown, and operation of the high-frequency vibration leveling mechanism are automatically controlled by the integrated control system. The steps include: the integrated control system receives real-time flatness data processed by the underwater camera; when the average elevation of the filling surface reaches the design elevation of the top surface of the leveling frame and the rate of change of the real-time flatness data is lower than the preset threshold, it automatically issues a start command; the high-frequency vibration leveling mechanism reciprocates along the preset track at a fixed speed; and when the integrated control system detects that the flatness deviation of the top surface of the pier does not exceed the preset difference threshold, it automatically stops the operation. The integrated control system is also equipped with a data recording and remote transmission module, which is used to record sensor data, control commands and equipment status information throughout the construction process, and transmit the sensor data, control commands and equipment status information to the shore-based monitoring center in real time through a satellite communication link; In step S2, pressure sensors are installed at the bottom of multiple legs of the leveling frame. After the top surface of the leveling frame is adjusted to the design elevation and kept horizontal, the integrated control system continuously monitors the pressure data at the bottom of each leg. If the pressure value of a single leg decreases by more than a first preset threshold, and the pressure value of one or more other legs increases by more than a second preset threshold, and the horizontal tilt angle data of the leveling frame fed back by the inertial navigation unit exceeds a third preset threshold, the integrated control system determines that the foundation under that leg has become unstable or eroded. The integrated control system issues an alarm and automatically suspends subsequent rock-throwing operations.

2. The high-precision leveling construction method for deep-water foundation trench crushed stone piers according to claim 1, characterized in that, Also includes: S5. After completing the slab throwing and leveling of a single crushed stone block, lift the leveling frame and the telescopic slab throwing funnel. S6. Use a workboat equipped with a multibeam echo sounder to scan and inspect the constructed crushed stone piers and conduct acceptance testing. S7. Repeat steps S1-S6 to construct the next pier until all piers are completed.

3. The high-precision leveling construction method for deep-water foundation trench crushed stone piers according to claim 1, characterized in that, Before lowering the leveling frame, a multibeam echo sounder system mounted on the workboat was used to conduct a full-coverage scan of the seabed trench to verify the elevation of the excavation surface.

4. The high-precision leveling construction method for deep-water foundation trench crushed stone piers according to claim 1, characterized in that, In the stone-throwing process, the crushed stone used has a particle size of 10-50mm and a continuous gradation.

5. The high-precision leveling construction method for deep-water foundation trench crushed stone piers according to claim 1, characterized in that, In step S3, the integrated control system dynamically adjusts the stone conveying rate based on the real-time flow velocity data at the funnel outlet monitored by the flow velocity sensor. Specifically, the integrated control system has a pre-stored correspondence table between flow velocity and conveying rate, which specifies the stone conveying rate value corresponding to different flow velocity ranges. The integrated control system queries the correspondence table and automatically adjusts the stone conveying rate to the corresponding target value according to the range of real-time flow velocity.

6. The high-precision leveling construction method for deep-water foundation trench crushed stone piers as described in claim 1, characterized in that, In step S3, the specific steps for acquiring real-time elevation and flatness data of the dumping surface include continuously acquiring underwater video streams during the dumping process using an underwater high-definition camera integrated on the retractable dumping funnel guide tube at a preset frame rate threshold, and transmitting the streams in real-time to the integrated control system via armored optical cable. The integrated control system then performs the following processing on the received image sequences: Each frame of the image undergoes processing including histogram equalization and physical model-based scattering effect elimination to enhance image contrast and edge features. Based on binocular vision or structured light 3D reconstruction algorithms, matching the same feature points in the left and right images, and generating a 3D point cloud dataset representing the current morphology of the dumping surface through triangulation. The 3D point cloud data is registered with the design coordinate system of the leveling frame, and a continuous digital elevation model is generated through gridding and interpolation algorithms. Based on the digital elevation model, the average elevation of the entire filling area and the flatness data used to characterize the degree of surface undulation are calculated in real time.

Images