Method for dynamically optimizing pile body construction sequence of water sinking pile

Abstract

The invention relates to the field of long-piled wharf engineering construction, and discloses an overwater pile sinking pile body construction sequence dynamic optimization method which comprises the following steps: collecting four-dimensional dynamic multi-source data in a construction process; constructing a real-time changing construction environment model based on the fused data; building a pile body construction sequence mathematical model according to the construction environment model and the wharf structure parameters; performing multi-objective optimization on the construction sequence mathematical model by adopting an intelligent algorithm, and dynamically generating a construction sequence and a ship moving path scheme; and a digital twinborn model of the pile driving barge and the pile body is constructed, the minimum distance between the barge body and the constructed pile body is predicted through a collision detection algorithm, and the construction sequence and the barge moving path are dynamically adjusted according to the detection result. According to the method, the construction safety, efficiency and global resource cooperation capability under complex working conditions are remarkably improved, and technical support is provided for intelligent construction of ocean engineering.

Description

Technical Field

[0001] The invention relates to the technical field of high-pile wharf engineering construction, and in particular to a method for dynamically optimizing the construction sequence of water-sunken pile bodies. Background Art

[0002] Traditional pile foundation construction sequence planning for high-piled wharfs relies heavily on manual experience and static design drawings. This decision-making process is limited by the complexities of the dynamic construction environment and the coupled influence of multiple factors. Existing methods typically employ fixed schedules or simple rules (such as sequential advancement of rows and columns). These methods struggle to integrate dynamic data such as tidal levels, soil parameters, and equipment status in real time, resulting in low construction efficiency and delayed risk control. Traditional static models are particularly incapable of adaptively adjusting construction paths in the face of sudden geological changes in soft soil, abnormal tides, or equipment failures, potentially leading to safety hazards such as pile displacement and ship collisions.

[0003] Furthermore, planning models driven by human experience lack quantitative trade-offs among multiple objectives, such as efficiency, cost, and safety, and often fall into local optimal solutions, resulting in extended construction periods and wasted resources. Current technologies are insufficient for data fusion and analysis in complex environments, and lack a closed-loop feedback mechanism. This results in construction plan adjustments relying on manual intervention, and the response speed and accuracy struggle to meet the high standards required for high-pile wharf projects.

[0004] Therefore, there is an urgent need for a pile foundation construction sequence planning method that integrates dynamic environment perception, multi-objective intelligent optimization and real-time feedback control to improve construction safety and efficiency under complex working conditions. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a dynamic optimization method for the construction sequence of underwater pile sinking. Under complex and changeable environments such as tides and geology, the dynamic coordinated optimization of the pile foundation construction sequence and the ship moving path is achieved, so as to simultaneously ensure construction efficiency, cost control and minimization of safety risks.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for dynamically optimizing the construction sequence of underwater pile bodies, comprising the following steps: Collecting four-dimensional dynamic multi-source data during the construction process, the four-dimensional dynamic multi-source data includes parameters of space, time, environment and resource dimensions, and fusing the data; Building a real-time changing construction environment model based on the fused data, the construction environment model integrates the time series prediction results and the continuous geological parameter distribution generated by spatial interpolation; According to the construction environment model and the wharf structure parameters, a mathematical model of the pile construction sequence is established, which includes geological condition constraints, pile arrangement constraints, and construction continuity constraints; An intelligent algorithm is used to perform multi-objective optimization on the mathematical model of the construction sequence, and a construction sequence and ship movement path plan is dynamically generated, wherein the optimization objectives include a weighted combination of efficiency, cost, and safety; Build a digital twin model of the piling vessel and pile body, predict the minimum distance between the hull and the constructed pile body through collision detection algorithm, and dynamically adjust the construction sequence and vessel movement path based on the detection results.

[0007] The present invention also provides a system for dynamically optimizing the construction sequence of underwater pile sinking, comprising: Data acquisition module, used to obtain four-dimensional dynamic multi-source data of the construction area in real time through multiple GPS positioning devices, geological detection equipment and environmental sensors; The construction sequence modeling module is used to build a mathematical model of the construction sequence based on pile position coordinates, geological condition parameters, and wharf structure parameters, and define a multi-objective optimization function for construction efficiency, cost, and safety risks; An intelligent algorithm optimization module uses a genetic algorithm to generate an initial plan for discrete sequence encoding of the pile construction sequence, an ant colony algorithm to model the ship movement path, and collaboratively optimizes the construction sequence and path plan based on a dynamic weight adjustment formula; A digital twin modeling module, which integrates ship parameters, designed pile parameters, and geographic data to construct a 3D dynamic twin model of the piling vessel, pile body, and construction environment; The collision detection module is used to calculate the minimum distance between the piling vessel and the constructed pile body through geometric algorithms, distinguish between square and round pile types, and output collision risk warnings; The dynamic planning module is used to trigger real-time adjustments to the construction sequence and ship movement path based on collision detection results, update the digital twin model based on the optimized plan, and output dynamic construction instructions.

[0008] The present invention provides a method for dynamically optimizing the construction sequence of underwater pile sinking. It has the following beneficial effects: 1. This invention utilizes the fusion of four-dimensional dynamic multi-source data and real-time environmental modeling to achieve second-level updates of construction parameters and the dynamic integration of multidimensional constraints. Compared to traditional static planning methods, it can accurately perceive sudden environmental changes such as tides and geology, dynamically adjust construction sequences and paths, and significantly reduce decision-making errors caused by data lag.

[0009] 2. Based on soil parameter interpolation and geological adaptability index classification, this invention constructs a quantitative mapping model between soil bearing capacity and construction risk. By setting hard boundaries for geological constraints (such as limiting continuous piling in soft soil areas), engineering risks such as soil liquefaction and pile displacement are effectively avoided, ensuring geological adaptability during the construction process.

[0010] 3. This invention utilizes a collaborative optimization framework combining a genetic algorithm and an ant colony algorithm, combined with a dynamic weight adjustment mechanism based on reinforcement learning, to overcome the limitations of traditional single-objective optimization. By adaptively prioritizing construction phases (e.g., prioritizing safety during positioning and efficiency during piling), it generates Pareto-optimal solutions under complex constraints, balancing global resource utilization with local risk control.

[0011] 4. Based on a digital twin model and a high-precision collision detection algorithm, this invention constructs a multi-dimensional dynamic perception system for the hull, pile, and environment. Through geometric distance calculation and risk heat map visualization, it enables millimeter-level obstacle avoidance adjustments to the ship's path, resolving the core pain points of delayed collision warnings and inefficient manual intervention in traditional construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Schematic diagram of the method flow of the present invention; Figure 2 Schematic diagram of the system architecture of the present invention.

[0013] Among them, 10. Data acquisition module; 20. Construction sequence modeling module; 30. Intelligent algorithm optimization module; 40. Digital twin modeling module; 50. Collision detection module; 60. Dynamic programming module. DETAILED DESCRIPTION

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0015] Please see the attached Figure 1 The present invention provides a dynamic optimization method for the construction sequence of underwater pile driving, which realizes dynamic planning of the entire construction process through multi-source data fusion, real-time environment modeling, intelligent algorithm optimization and digital twin simulation.

[0016] like Figure 1 As shown, the method for dynamically optimizing the construction sequence of underwater pile bodies may include the following steps: S1. Collect four-dimensional dynamic multi-source data during the construction process; S2. Build a real-time changing construction environment model based on the fused data; S3. Establish a mathematical model of the construction sequence based on the construction environment model and the wharf structure parameters; S4. Use intelligent algorithms to perform multi-objective optimization on the mathematical model of the construction sequence and dynamically generate the construction sequence and ship movement path plan; S5. Build a digital twin model of the piling vessel and pile body, predict the minimum distance between the hull and the constructed pile body through the collision detection algorithm, and dynamically adjust the construction sequence and vessel movement path based on the detection results.

[0017] The following is a detailed description of each step in the method of the present invention, which comprehensively explains the specific implementation principles, technical details and processes of each step.

[0018] Regarding step S1, in this embodiment, the collection and fusion processing of four-dimensional dynamic multi-source data is achieved by the following method: Integrate multi-source sensors and data interfaces to obtain dynamic parameters of space, time, environment and resource dimensions during the construction process in real time.

[0019] Spatial dimension data includes but is not limited to the ship parameters of the pile-driving vessel (such as hull length, width, and draft), the three-dimensional coordinates (X, Y, Z) and attitude parameters (pitch angle, roll angle) of the hull collected in real time by multiple high-precision GPS positioning devices, and the designed pile position coordinates and elevation data extracted from the dock design drawings.

[0020] Time dimension data covers the time series division of the construction period (such as tidal cycles, meteorological change periods) and historical environmental data (such as tide tables and wind speed history records in the past three years), which are used to establish a time correlation analysis model.

[0021] Environmental dimension data is collected in real time through ship-borne weather stations, buoy-type water level gauges and geological exploration equipment, including meteorological monitoring data (wind speed, wave height, visibility), tidal water level data (current water level and predicted water level change curve), geological scanning data (standard penetration number), , Static penetration cone tip resistance , soil cohesion , internal friction angle and groundwater depth ).

[0022] Resource dimension data is obtained through the equipment status monitoring system, including equipment operating parameters such as the pile-driving vessel's fuel consumption rate, hydraulic system pressure, and pile hammer striking frequency, as well as the pile inventory, replenishment time window, and barge positioning coordinates of the pile-transporting vessel.

[0023] The above multi-source heterogeneous data is integrated and processed, specifically including: (1) Data deduplication and cleaning: Redundancy of repeatedly collected data is eliminated based on spatiotemporal tags. For example, the data with the highest accuracy is retained from multiple geological scans of the same pile coordinate point. (2) Missing value filling: For the problem of missing geological parameters in some areas, the neighboring point interpolation method or the classification mean based on the soil layer type is used to fill in the missing values. For example, when a certain exploration point When a value is missing, the average value of the same soil layer is calculated based on the soil layer category (such as clay, sand) to fill it in; (3) Spatiotemporal alignment of multi-source data: The Kalman filter algorithm is used to align the time series of data sampled at different frequencies. For example, high-frequency GPS coordinate data (sampling rate 10 Hz) and low-frequency tidal data (updated once an hour) are unified to the same timestamp, eliminating the timing deviation and generating a temporally and spatially consistent data stream.

[0024] Furthermore, geological parameters in the environmental dimension data are sampled at discrete points using sonar detectors and geological radar. The sampling interval is dynamically adjusted based on geological complexity, for example, increasing the density of sampling points in areas of sudden changes in soil layers (such as the interface between soft and hard soil layers). Equipment status parameters in the resource dimension data are transmitted to the central processing unit in real time via the CAN bus and IoT sensors. These parameters are then linked to the planned data in the construction progress management system to monitor the matching of resource supply and consumption in real time.

[0025] The Kalman filter algorithm preferably uses a state-space model during the data alignment process, using GPS coordinate data as observation variables and tidal and meteorological data as state variables. Time synchronization of multi-source data is achieved through a prediction-update cycle. Specifically, the prediction phase establishes state transition equations based on historical data, while the update phase corrects prediction errors using observation data, ultimately outputting a multidimensional dataset with aligned timestamps.

[0026] Through the above-mentioned method, this embodiment fully realizes the efficient collection, cleaning and fusion of four-dimensional dynamic multi-source data, providing a high-precision and high-reliability data foundation for subsequent construction environment modeling and intelligent optimization.

[0027] Regarding step S2, in this embodiment, the construction of the real-time changing construction environment model is achieved by the following methods: Integrate time series prediction and spatial interpolation algorithms to generate a multidimensional dynamic environmental model of the construction area.

[0028] Specifically, time series prediction is based on historical meteorological data and tidal data. A long short-term memory neural network (LSTM) is used to build a prediction model. Environmental parameters within the historical time window (such as wind speed, wave height, and water level changes) are input to output the dynamic change trend of the future construction period.

[0029] The LSTM network captures the long-term dependencies of time series data through the forget gate, input gate, and output gate mechanisms. For example, it can jointly model periodic tidal changes and sudden meteorological fluctuations to ensure that the prediction results take into account both regularity and the ability to respond to abnormal events.

[0030] For the continuous spatial distribution modeling of geological parameters, the spatial interpolation algorithm is based on the soil parameters of discrete exploration points (such as standard penetration number , cone tip resistance ) for interpolation calculation. Preferably, in areas where the spatial correlation of soil parameters is strong (such as uniform soft clay layer), the Kriging interpolation method is used to fit the spatial autocorrelation structure based on the semivariogram, which is expressed as: , in, is the lag distance, is the lag distance The number of point pairs inside, For location For areas with sparse exploration points or sudden geological changes, the inverse distance weighted interpolation method (IDW) is used to calculate the target point parameter value by weighting the inverse distance: , in, Target point With known points The Euclidean distance of is the weight index (preferably ranging from 1.5 to 2.5).

[0031] The dynamic integration of the construction environment model is achieved through: Combine the time series prediction results (such as the wind speed change curve for the next 6 hours) with the continuous geological parameter distribution generated by spatial interpolation (such as The grid map is superimposed on the three-dimensional geographic model to form a four-dimensional (space + time) dynamic environmental field. For example, during the period of rising tide level, the model automatically updates the calculation boundary conditions of underwater soil parameters (such as groundwater level). dynamic adjustment with tidal changes) and linked to the real-time calculation of geological suitability indicators.

[0032] Preferably, the dynamic environmental model adopts a layered architecture design, with the bottom layer being geospatial data (terrain elevation, pile position coordinates), the middle layer being real-time environmental parameters (wind speed vector field, tidal water level surface), and the top layer being dynamic prediction and interpolation results (future meteorological trends, soil layer parameter distribution).

[0033] The model update mechanism combines event-driven and timed refresh: when the deviation between the newly collected geological data and the prediction exceeds a preset threshold (such as the wind speed prediction error >15%), the local model re-interpolation is triggered; otherwise, the global model is refreshed once at a fixed period (such as every 30 minutes).

[0034] Through the above-mentioned method, the construction environment model constructed in this embodiment has the capabilities of multi-dimensional data fusion, dynamic update and risk warning, providing a high-precision environmental perception basis for pile construction sequence and path planning.

[0035] Regarding step S3, in this embodiment, the establishment of the mathematical model of the pile construction sequence is achieved by the following method: The construction sequence modeling is based on the dynamic environment model and the wharf design parameters. A mathematical model is constructed that includes geological constraints, pile layout constraints, and construction continuity constraints, and a multi-objective optimization function is defined. Specifically, the following steps are used to implement variable definition, constraint modeling, and objective function design: The decision variable definition includes the pile set (in Indicates the coordinates of the piles), construction sequence variables (Indicates the pile position number for each construction), ship moving path variable (Indicates the Starting point of the next ship transfer and the end point coordinates), ship moving batch variables (Indicates the Whether to use the first construction step The position after the ship is moved. indicates use), and construction time variables (Indicates the The construction sequence variable must satisfy the uniqueness constraint. , ensuring that each pile position is constructed only once.

[0036] In this embodiment, the geological condition constraints are based on the construction environment model constructed in step S2 to extract the geological parameters of the construction area (soil layer distribution, bearing capacity, groundwater level, etc.). The constraints are quantified as follows: Define the geological adaptability index for pile construction. The geological adaptability index is based on the quantification of geological parameters. The specific requirements include: Soil layer distribution: Penetration resistance thresholds are set for different soil layers. In soft soil areas, the number of consecutive pile driving times needs to be limited to avoid soil disturbance that may cause pile displacement.

[0037]

[0038] Determination of high bearing capacity area: The core indicators for determining high bearing capacity areas include: Standard Penetration Test (SPT), Corrected Blow Count ;Cone penetration test (CPT), cone tip resistance ; Geotechnical test: cohesion , friction angle .

[0039] Geological adaptability index is a comprehensive geological parameter. The higher it is, the more suitable it is for piling: , in: is the standard penetration test blow number (dimensionless), reflecting the density and shear strength of the soil layer; is the cone tip resistance of static penetration (kPa), which represents the soil strength; is the soil cohesion (kPa), which measures the bonding force between soil particles; is the internal friction angle of soil (°), reflecting the soil's ability to resist shear; is the depth of groundwater level (m), which affects soil stability; is the critical groundwater level (fixed value), exceeding which may cause soil liquefaction or loss of bearing capacity; is the weight coefficient, the default value is (0.3, 0.3, 0.2, 0.2), satisfying ; Dominant (0.3 each), reflecting the core influence of standard penetration blows and cone tip resistance on geological adaptability; (0.2) Comprehensive cohesion and internal friction angle, reflecting the shear strength of soil; (0.2) Penalize high groundwater levels and lower the index value.

[0040] In this embodiment, according to the engineering geological survey results and design requirements, the preset grading standards.

[0041]

[0042] In this embodiment, the wharf structure constraint sets construction rules according to the pile arrangement in the design drawings.

[0043] Structural parameter extraction: According to the wharf design drawings, obtain parameters such as pile arrangement (determinant, pile group), pile spacing, and pile top elevation.

[0044] The constraints of pile arrangement, pile spacing and pile top elevation on the construction sequence: (1) Pile arrangement (determinant / group pile) Determinant: Piles are arranged in rows or columns, and construction proceeds in a row-first or column-first order. For a three-column, five-row determinant layout, prioritize all piles in the first row, followed by the second and third rows. This reduces the need for frequent lateral movement of the piling barge and improves construction efficiency.

[0045] Pile groups: Clustered piles (e.g., plum blossom or rectangular pile groups) require either a "grouped construction" or "symmetrical construction" strategy. If the pile group is divided into three groups, A, B, and C, each containing five piles, they should be constructed in the order A→B→C, or symmetrically (e.g., A1→A3→A5, then B1→B3→B5). This minimizes soil uplift or lateral displacement caused by the pile group effect and ensures uniform settlement of the pile foundation.

[0046] (2) Pile spacing constraints Minimum safety distance in pile foundation engineering It is a key parameter to ensure that piles avoid mutual interference during construction and control soil deformation and vibration. Based on pile foundation engineering experience, the minimum safe distance is usually calculated using the following formula: , in: is the pile diameter (m), directly take the designed pile diameter (such as 0.5m, 1.0m, etc.); is the cone tip resistance of the static penetration test (kPa), obtained through the static penetration test (CPT) or the standard penetration test (SPT); is the soil density (kN / m 3 ), provided by geological survey report.

[0047] (3) Pile top elevation constraint If the difference in pile top elevation exceeds the design allowable value (e.g., Δh > 1.5 m), construction must be carried out in layers, from deep to shallow. This prevents the pile-driving vessel from becoming unstable due to an excessively long cantilever and reduces the impact of soil rebound on the installed piles. For example, in a dock area, where the pile top elevations are -10 m, -8 m, and -5 m, respectively, construction of the -10 m pile group is required first, followed by construction down to the -5 m pile group.

[0048] If the pile top elevation corresponds to different structural layers (such as the cap layer and the beam-slab layer), ensure that the bottom layer of piles is constructed first, forming a stable support before constructing the upper layer of piles. For example, if the pile cap elevation is -3m and the beam-slab elevation is +2m, all piles at -3m must be constructed first, followed by the +2m piles.

[0049] Based on the constraints of the aforementioned wharf structures (e.g., continuous beams, independent piers) and geological conditions, wharf pile foundation construction adopts a "block-by-block" approach, with each block divided into different types. The entire construction area is divided into several independent construction zones. For example, a wharf measuring 500 meters long and 30 meters wide can be divided into 10 zones measuring 50 meters by 30 meters. Areas with complex geological conditions (e.g., soft soil) are designated separately to avoid overlapping construction with other zones.

[0050] In this embodiment, the construction sequence within the block is regular, and the form is redivided according to the following steps, forming an efficient flow construction process of "continuous supply of pile foundation and continuous driving of whole barges".

[0051] (1) Prioritize the construction of key piles If it is a continuous beam structure, give priority to constructing the piles supporting the main beam (such as at the mid-span or support point); If it is a pile group foundation, priority should be given to constructing the center pile of the pile group to reduce soil disturbance.

[0052] (2) Concentrated construction of similar piles Piles of the same type and elevation within the same block are constructed together to reduce the time required to adjust the piling vessel equipment. For example, all square piles in the deep area (-15m) are constructed first, followed by round piles in the shallow area (-10m).

[0053] (3) Symmetrical construction reduces stress Symmetrically arranged piles (such as symmetrical piles on both sides of a pier) should be constructed in a symmetrical sequence to avoid structural overloading; example: A1→A3→A5→B1→B3→B5.

[0054] In this embodiment, the piling vessel operating space constraint defines the accessible area and restricted area through equipment model parameters, with the goal of ensuring construction safety and efficiency, determining the operating range of the piling vessel, and defining the accessibility conditions for pile construction.

[0055] The workspace constraints are as follows: Based on the established piling vessel equipment model, the size of the piling vessel, the anchoring position, the area covered by the cable points, the minimum turning radius and obstacle avoidance space of the vessel movement path are determined, and the entire construction area is divided into blocks to ensure the construction continuity of the piling vessel.

[0056] Key parameters of the pile driving vessel equipment model The operation capacity of a pile driving vessel is determined by its physical parameters and dynamic performance, mainly including the hull size (length L, width W, draft D), operation range (anchor point coordinates, pile hammer coverage radius ) (2) Constraints on accessible areas of pile-driving vessels The reachable area of ​​a pile-driving vessel, i.e., the operating range of a pile-driving vessel, is determined by its anchoring point and is usually a polygonal or circular area: , in: is the anchor point coordinate set; The radius is centered at the anchor point. Circular coverage area; The number of anchoring points (usually 2 to 4).

[0057] Each pile must be within the coverage of the pile driving vessel after a certain ship movement: Make

[0058] in: is the index of the construction step, The value range of is 1 to n (the total number of piles); is the construction sequence variable, representing the The number of the pile being constructed in each construction step; For the The coordinates of the pile being constructed in each construction step; An index for positioning the moving ship; For the The coordinates of the end point of the second ship movement positioning, that is, the position of the pile driving ship during operation; For pile driving vessels located The scope of operations that the premises can cover; is a universal quantifier, indicating that this constraint must be applied to all construction steps Established.

[0059] is an existential quantifier, indicating that for each step , there is at least one ship slot Can meet the conditions.

[0060] (3) Restrictions on the restricted area for pile-driving vessels To avoid collision with the hull, a safety restricted area must be set up around the constructed piles.

[0061] , , in: is the coordinate of the constructed pile; is the minimum safe distance; is the number of constructed piles; is the pile diameter (m); is the pile end bearing capacity (kPa), obtained by the cone penetration test (CPT); is the soil density (kN / m 3 ); is the vibration influence coefficient; is the depth of groundwater level (m).

[0062] In this embodiment, the construction continuity constraints include the continuity of the ship moving path, the tidal time window restriction, and the time accumulation rule, as follows: (1) Ship moving path continuity constraint The end point of the ship moving path is the starting point of the next ship moving: ,in The number of ship transfers.

[0063] (2) Tidal time window constraints The construction of all piles must begin within the permitted tidal range: , ,in: For the The construction start time of the root pile.

[0064] Represents a continuous tidal time window during which construction is permitted.

[0065] is the earliest allowed start time for the window.

[0066] is the latest allowed start time for this window.

[0067] (3) Time constraints Construction time and ship moving time are accumulated alternately: , in, For pile construction time, The time for moving the ship.

[0068] In this embodiment, the multi-objective optimization function integrates efficiency, cost, and safety goals, and integrates efficiency (total time), cost (total expense), and safety (total risk) into a weighted single objective function: , in, is the total time, is the total cost, For budget costs, is the total risk, is the risk threshold, is the weight coefficient.

[0069] Decompose the objective function: (1) Total time : , in: , As a function, calculate the Starting point of the next ship transfer and the end point The straight-line distance between is the moving speed of the pile driving vessel; and Indicates the The starting and ending coordinates of the ship movement.

[0070] (2) Total cost : , in: For pile construction costs, , The cost of moving the pile-driving ship per unit distance.

[0071] (3) Total risk : , in: is the geological risk (such as soil liquefaction probability, ranging from 0 to 1); is the environmental risk (e.g. tidal influence coefficient, ranging from 0 to 1); is the operational risk (such as collision probability, ranging from 0 to 1).

[0072] Based on the three types of first-level indicators of total risk decomposition, the second-level indicators are refined:

[0073] Each secondary indicator is given a fuzzy score (0-1, 0 for no risk, 1 for extremely high risk). Combined with historical data, the probability distribution of each indicator (such as normal distribution, triangular distribution) is determined, and the joint probability is calculated: , in: For the The probability of a risk event occurring; is the total number of risk events.

[0074] Preferably, the weight coefficient 、 、 Dynamic adjustments are made according to the construction stage (e.g., safety is prioritized during the positioning stage, while efficiency is prioritized during the piling stage).

[0075] The model minimizes the weighted total time, cost and risk by optimizing the construction sequence, ship movement path and time arrangement while satisfying the constraints of uniqueness, safety distance, equipment accessibility and tidal window.

[0076] Through the above method, the mathematical model constructed in this embodiment completely covers geological adaptability, structural form, working space and continuity constraints, and realizes dynamic planning of construction sequence and path through multi-objective optimization, ensuring the feasibility of the technical solution and scientific decision-making.

[0077] In step S4, in this embodiment, the intelligent algorithm multi-objective optimization is implemented through a collaborative framework of genetic algorithm, ant colony algorithm and reinforcement learning to dynamically generate the construction sequence and ship movement path plan. The specific implementation process is as follows: The collaborative framework combining intelligent algorithms and reinforcement learning includes: Intelligent algorithms are responsible for generating initial feasible solutions, which serve as the initial policy library for reinforcement learning. Reinforcement learning is responsible for fine-tuning, and the feedback data provides a feasible solution space for multi-objective optimization. Preferably, reinforcement learning uses a deep Q-network (DQN) and proximal policy optimization (PPO) to respond to environmental changes in real time.

[0078] In this embodiment, the initial scheme generation and screening is based on the mathematical model established in step S3, and a genetic algorithm is used to perform discrete sequence encoding on the construction sequence variables to generate a pile position number sequence. The specific process is as follows: 1) Design a coding method to encode the construction sequence variable into a discrete sequence (such as the pile position number sequence A→B→C→D); the ship moving path is a set of path segments, , indicating the Starting point of the next ship transfer and end point coordinate.

[0079] 2) Avoid risk areas according to geological constraints ( ) must be excluded, and the number of consecutive pile driving times is limited (e.g. ≤ 3 piles in soft soil areas); Among the constraints on the wharf structure, row-column layouts must be arranged according to row / column priority, pile groups must be constructed symmetrically or in groups, and pile spacing and pile top elevations must be constructed in layers. As for the piling vessel's operating space constraints, the pile position must be within the piling vessel's operating coverage (polygonal or circular area), and the vessel's moving path must meet a safe distance (restricted area around the constructed pile). Randomly generate an initial construction sequence population that satisfies the constraints.

[0080] 3) Eliminate solutions that violate construction continuity constraints (e.g., discontinuous ship movement paths) or equipment capacity constraints (e.g., overloaded lifting). The selected feasible solutions are stored in the strategy library as the initial input for reinforcement learning.

[0081] Furthermore, this embodiment transforms construction sequence optimization into a multi-objective combination optimization problem based on historical construction data, environmental prediction models, and equipment parameters. The objective functions include efficiency, safety, energy consumption, etc., and reversely screens ineffective solutions from feasible solutions. The specific process is as follows: 1) Based on the objective function in step S3 , define the weighted combination of efficiency (time), cost, and safety (risk), and dynamically adjust the weights , , .

[0082] 2) The optimization target weights are dynamically adjusted according to the construction stage. Key targets include efficiency weight, safety weight, and energy consumption weight.

[0083] 3) The construction process is divided into phases: positioning, piling, and ship transfer. Each phase has different priorities. The positioning phase, which prioritizes ship stability, has the highest safety weighting; the piling phase has a higher efficiency weighting; and the ship transfer phase, which prioritizes avoiding completed piles and environmental interference, has a higher energy consumption weighting. Based on the characteristics of the construction phase, the initial weightings for each objective are as follows:

[0084] 4) Dynamic weight adjustment is a process of dynamically optimizing weights through mathematical models based on real-time data and environmental changes.

[0085] The dynamic initial weight adjustment method is as follows: , , , in, The foundation weight is preset during the construction phase; Indicates the impact of real-time environmental changes; Indicates device status feedback; Indicates the Real-time value of each device status; is with The corresponding design upper limit, rated capacity or safety threshold of the equipment; It is Real-time measurements of environmental factors; is with Preset thresholds, reference upper limits or safety critical values ​​of corresponding environmental factors; is the weight adjustment coefficient.

[0086] 5) Based on the objective function and constraints (such as geological adaptability, wharf structure, and piling vessel operation space), reversely screen the generated invalid solutions.

[0087] In this embodiment, the genetic algorithm optimizes the construction sequence by exploring the solution space through crossover and mutation, generating new solutions, avoiding falling into local optimality, and eliminating invalid solutions that violate construction continuity or equipment capacity. The specific process is as follows: 1) Pile sequence coding crossover: select the pile exchange sequence based on the initial plan (for example, the parent generations [A, B, D] and [C, E, F] generate the child generations [A, B, C] and [D, E, F]), retaining some pile construction sequences to avoid violating geological constraints.

[0088] 2) Pile sequence coding mutation: According to the initial plan, the pile mutation sequence is selected and the pile positions are randomly exchanged or inserted (such as [A,B,D]→[B,A,D] or generating offspring [A,B,C] and [A,C,B]).

[0089] In this embodiment, the ship moving path is optimized based on the ant colony algorithm, and the optimal ship moving path set is output to ensure the shortest path and the lowest safety risk. The specific process is as follows: 1) The construction area is abstracted into a graph, with the nodes being piles and the edges being ship moving paths. The edge weight is defined as the combined value of the path length and the safety factor, which is calculated through collision monitoring. The minimum distance between the constructed pile and the path through real-time monitoring and safety thresholds Calculate. The formula is:

[0090] 2) Initialize the information density of all paths ,ensuring that all paths have equal probability of being selected in the initial exploration phase of the algorithm,,avoiding prior bias.,The higher the pheromone concentration on a path, the more times the path is selected by ants,,and the more likely it is to be better.

[0091] 3) The pheromone of the unselected path gradually decreases ( , is the volatility coefficient, usually set to ρ = 0.5). The selected path pheromone increases ( , is the pheromone increment of the new path). Through volatilization and enhancement, the algorithm gradually focuses on high-quality paths and eventually converges to the global optimum.

[0092] , in, For path segments length.

[0093] 4) Heuristic information is a heuristic function that evaluates the quality of a path and is used to guide ants to choose a more likely path. Since the optimization goal is to minimize the total path length, the inverse of the path distance can directly reflect the quality of the path. The shorter the distance, the better the heuristic value. The larger the value, the higher the probability that the ant will choose this path. This simplifies the evaluation of complex environments and enables the algorithm to quickly focus on potential high-quality paths.

[0094] , in: For path segments length.

[0095] 5) An ant colony algorithm optimizes the boat movement paths and outputs an optimal set of paths, determined by both pheromone concentration and heuristic information. Pheromone concentration records the historical quality of paths, while heuristic functions based on path attributes (such as distance) are dynamically updated by evaporating and enhancing them, guiding ants to quickly discover potentially optimal paths. This combination of the two allows the algorithm to strike a balance between exploring new paths and leveraging historical experience, ultimately finding the global optimal solution.

[0096] , in: is the importance weight of pheromone. ; is the importance weight of the heuristic information. In this embodiment ; Indicates the sum of all optional paths.

[0097] In this embodiment, the generated initial solution provides a feasible solution space for multi-objective optimization, while constraints are used to filter out ineffective solutions. Genetic algorithms and ant colony algorithms optimize the construction sequence and ship movement path, respectively. Their outputs serve as dynamic input parameters for multi-objective optimization, forming a closed-loop iteration. Simultaneously, the multi-objective optimization objective function (e.g., total time, cost) drives the optimization direction of the genetic and ant colony algorithms. Dynamic optimization is achieved by adjusting parameters of the overall process (e.g., crossover probability in the genetic algorithm) through real-time feedback (e.g., equipment status, environmental changes).

[0098] In step S5, dynamic monitoring and feedback optimization of the construction process are achieved through the collaboration of the digital twin model and the collision detection algorithm, specifically through the following methods: In this embodiment, the digital twin model is constructed based on the four-dimensional dynamic multi-source data in step S1 to establish a high-precision model of the piling vessel, pile body and construction environment.

[0099] The hull model is based on the ship parameters (length L, width W, draft D) and real-time GPS positioning data (three-dimensional coordinates ) and the positioning error is eliminated by spatial differential calibration of three GPS devices.

[0100] The pile model is based on the design parameters (pile pitch angle , pile length , pile diameter , slope ratio ) is constructed, and its three-dimensional space equation is expressed as:

[0101] in, is the three-dimensional coordinate of a known reference point on the pile.

[0102] The environmental model integrates terrain elevation data, real-time weather conditions (wind speed , high waves ) and tidal water level prediction curves to form a three-dimensional scene that is consistent in time and space.

[0103] In this embodiment, collision detection and minimum distance calculation use differentiated geometric algorithms based on the pile type (round pile or square pile). The specific process is as follows: 1) Determine whether it is a round pile or a square pile based on the designed pile parameters.

[0104] 2) If it is determined to be a square pile, the vector geometry algorithm is used to calculate the minimum distance between the hull and the edge of the square pile. The pre-judgment distance is divided into the distance between edges and the distance between vertices and edges. The minimum distance for:

[0105] in, is the calculated shortest distance between the hull and the pile; It is a general distance calculation function that represents the shortest Euclidean distance between two geometric objects (points or line segments) as input. In this formula, it is used in three cases: Represents a specific edge of the hull A specific side with a square stake The shortest distance between.

[0106] Represents a specific vertex of the hull A specific side with a square stake The shortest distance between.

[0107] Represents a specific vertex of the square stake A specific side of the hull The shortest distance between.

[0108] and Respectively represent the hull Edge and vertices.

[0109] Edge Set ; Vertex Set ; and Respectively represent the first Edge and vertices.

[0110] Edge Set ; Vertex Set

[0111] is the minimum value function.

[0112] 3) If it is determined to be a circular pile, the minimum distance from the pile to the hull is determined by the distance from the pile center to the edge set and vertex set of the hull, and the geometric algorithm is used to express the minimum distance from the pile to the hull using the distance from point to line and the distance from point to point. for: , in, Represents the hull edge, edge set , Represents the hull vertex, vertex set ; Indicates the coordinates of the pile center.

[0113] represents the distance from the pile center to the hull edge set, Indicates the distance from the pile center to the hull vertex set.

[0114] 4) If Less than the safety threshold , triggering a collision warning and replanning the ship moving path based on the ant colony algorithm, updating the construction status of the pile in the digital twin model, and iteratively optimizing the construction sequence and path plan.

[0115] Through the above steps, the minimum distance between the pile-driving vessel hull and the completed pile is determined, dynamic planning is triggered, and the vessel movement path is adjusted.

[0116] In general, the present invention collects multi-source data, constructs a dynamic environmental model, establishes a multi-objective function mathematical model, uses intelligent algorithms to establish a dynamic weight adjustment mechanism, optimizes the construction sequence, and combines spatial geometry algorithms and collision warning and monitoring technologies to further optimize the ship moving path.

[0117] The system for dynamically optimizing the construction sequence of underwater pile bodies described below and the method for dynamically optimizing the construction sequence of underwater pile bodies described above can be used for reference in correspondence with each other.

[0118] Please see the attached Figure 2 The present invention also provides a system for dynamically optimizing the construction sequence of underwater pile sinking piles, comprising: The data acquisition module 10 is used to obtain four-dimensional dynamic multi-source data of the construction area in real time through multiple GPS positioning devices, geological detection equipment and environmental sensors; The construction sequence modeling module 20 is used to construct a mathematical model of the construction sequence based on the pile position coordinates, geological condition parameters and wharf structure parameters, and define a multi-objective optimization function for construction efficiency, cost and safety risk; Intelligent algorithm optimization module 30 is used to generate an initial plan by discretely encoding the pile construction sequence using a genetic algorithm, model the ship movement path using an ant colony algorithm, and collaboratively optimize the construction sequence and path plan based on a dynamic weight adjustment formula; Digital twin modeling module 40, used to integrate ship parameters, designed pile parameters and geographic data to build a three-dimensional dynamic twin model of the piling ship, pile body and construction environment; Collision detection module 50, used to calculate the minimum distance between the piling vessel and the constructed pile body through geometric algorithms, distinguish between square piles and round piles and output a collision risk warning; The dynamic planning module 60 is used to trigger real-time adjustment of the construction sequence and ship movement path according to the collision detection results, update the digital twin model based on the optimized plan, and output dynamic construction instructions.

[0119] The system of this embodiment can be used to execute the above method embodiments, and its principles and technical effects are similar, so they will not be repeated here.

[0120] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamically optimizing the construction sequence of underwater pile sinking, characterized in that: The following steps are involved: Collecting four-dimensional dynamic multi-source data during the construction process, the four-dimensional dynamic multi-source data including parameters of space, time, environment and resource dimensions, and fusing the four-dimensional dynamic multi-source data; Building a real-time changing construction environment model based on the fused data, the construction environment model integrates the time series prediction results and the continuous geological parameter distribution generated by spatial interpolation; According to the construction environment model and the wharf structure parameters, a mathematical model of the pile construction sequence is established, which includes geological condition constraints, pile arrangement constraints, and construction continuity constraints; An intelligent algorithm is used to perform multi-objective optimization on the mathematical model of the construction sequence, dynamically generating a construction sequence and ship movement path plan, with the optimization objectives including a weighted combination of efficiency, cost, and safety; Build a digital twin model of the piling vessel and pile body, predict the minimum distance between the hull and the constructed pile body through collision detection algorithm, and dynamically adjust the construction sequence and vessel movement path based on the detection results.

2. The method for dynamic optimization of the construction sequence of underwater pile sinking according to claim 1, characterized in that: The four-dimensional dynamic multi-source data includes: Spatial dimension data, including the piling vessel's ship parameters, real-time position coordinates, and designed pile position coordinates and elevations; Time dimension data, including time series of the construction period and historical environmental data; Environmental dimension data, including meteorological monitoring data, tidal water level data and geological scanning data; Resource dimension data, including equipment status parameters of pile-driving vessels and supply status parameters of pile-carrying vessels; The fusion processing includes data deduplication, missing value filling and multi-source data alignment based on Kalman filtering.

3. The method for dynamic optimization of the construction sequence of underwater pile sinking according to claim 1, characterized in that: The step of constructing a real-time changing construction environment model based on the fused data includes: Use long-short-term memory neural networks to perform time series forecasts on historical meteorological and tidal data, and output dynamic trends of environmental parameters during the future construction period; Use Kriging interpolation or inverse distance weighted interpolation to spatially interpolate soil layer parameters at discrete geological exploration points to generate a continuous geological distribution model for the construction area; The time series prediction results and spatial interpolation results are integrated to build a dynamically updated construction environment model, and the model parameters are refreshed in real time with the newly collected data.

4. The method for dynamic optimization of the construction sequence of underwater pile sinking according to claim 1, characterized in that: The step of establishing a mathematical model of pile construction sequence including geological condition constraints, pile arrangement constraints, and construction continuity constraints based on the construction environment model and wharf structural parameters includes: Define the construction sequence variables, ship moving path variables and time variables, and constrain each pile to be constructed only once; The geological condition constraints set the upper limit of the continuous piling times and the geological adaptability index based on the soil layer parameters in the real-time environmental model. The index is calculated by weighting the standard penetration blows, cone tip resistance, cohesion, internal friction angle and groundwater level parameters. Pile arrangement constraints define the construction sequence rules of rows or groups of piles according to the wharf design drawings, including row / column priority sorting or symmetrical grouping construction, and the pile spacing must meet the minimum safety distance; Construction continuity constraints include the start-end continuity of the ship moving path, tidal time window restrictions, and the pile body being located within the operation coverage area of ​​the pile driving ship.

5. The method for dynamic optimization of the construction sequence of underwater pile bodies according to claim 4, characterized in that: The calculation formula of the geological adaptability index is: ， in, Corrected number of blows for standard penetration test, is the cone tip resistance of static penetration test, is the soil cohesion, is the internal friction angle, is the depth of groundwater level, is the maximum design value of the corresponding parameter, is the critical groundwater level, is the normalized weight coefficient.

6. The method for dynamic optimization of the construction sequence of underwater pile sinking according to claim 4, characterized in that: The objective function of the multi-objective optimization of the mathematical model of pile construction sequence is a weighted combination of efficiency, cost and safety, and its expression is: ， in, is the total time, is the total cost, For budget cost, is the total risk, is the risk threshold, is the weight coefficient.

7. The method for dynamically optimizing the construction sequence of underwater piles according to claim 6, characterized in that: The step of using an intelligent algorithm to perform multi-objective optimization on the construction sequence mathematical model to dynamically generate a construction sequence and a ship moving path plan includes: A collaborative optimization framework of genetic algorithm and ant colony algorithm is adopted. The genetic algorithm discretizes the sequence of pile construction, generates a pile position number sequence, and generates an initial feasible solution based on geological condition constraints, construction continuity constraints, and pile layout constraints. Dynamically adjust weight coefficients based on the objective function , reversely screen the invalid solutions among the initial feasible solutions; The weight coefficient of the objective function Dynamic adjustment according to construction stage: Positioning stage: security weight Raise priority based on device stability feedback; Pile driving stage: efficiency weight Dynamically increase according to the progress of pile construction; Ship moving stage: energy consumption weight Combine fuel consumption rate and path length optimization; The weight is dynamically adjusted through the formula: ， in, is the stage basic weight, is the environmental impact factor, For the The maximum allowable change of class environment parameters, Provides feedback on device status; is the weight adjustment coefficient; New plans are generated through crossover and mutation of genetic algorithms, invalid solutions that violate construction continuity or equipment capacity are eliminated, and the optimized construction sequence is fed back to the ant colony algorithm to iteratively update the ship moving path.

8. The method for dynamically optimizing the construction sequence of underwater piles according to claim 7, characterized in that: The step of iteratively updating the ship moving path using the ant colony algorithm includes: The construction area is modeled as a graph structure, with nodes representing pile position coordinates and edges representing ship movement path segments. , the edge weight is defined as ,in, Indicates the Starting point of the second ship transfer and end point coordinate, For path segments The Euclidean distance of The minimum distance between the constructed pile and the path through real-time monitoring and safety thresholds Calculation formula is: ， Initialize the pheromone concentration of all paths , dynamically updated through pheromone volatilization and enhancement mechanism: ， in, is the pheromone volatility coefficient; The probability of an ant choosing a path is determined by the pheromone concentration and heuristic information Jointly decided, the calculation formula is: ， in, is the pheromone weight coefficient, is the heuristic information weight coefficient; By iteratively updating the pheromone concentration and path selection probability, the global optimal ship moving path set is output.

9. The method for dynamically optimizing the construction sequence of underwater piles according to claim 1, characterized in that: The steps of constructing a digital twin model of the piling vessel and the pile, predicting the minimum distance between the vessel and the constructed pile using a collision detection algorithm, and dynamically adjusting the construction sequence and vessel movement path based on the detection results include: The 3D hull model of the pile driving ship is constructed based on the ship parameters and the real-time coordinate signals of multiple GPS positioning devices. Represents the four edges of the hull, the vertex set Represent the four vertices of the hull; According to the designed pile parameters, the three-dimensional space equation of the pile body is established to distinguish between round piles and square piles. Represents the four sides of the square pile, the vertex set The four vertices of the square pile are represented by and the coordinates of the center point of the circular pile are ; When the pile is a square pile, calculate the hull edge set Side set with square pile Minimum distance between , the formula is: ， in, is the shortest distance between the hull edge set and the pile edge set, is the vertical distance from the hull vertex set to the square pile edge set, is the vertical distance from the pile vertex set to the hull edge set; When the pile is a circular pile, calculate the center point of the circular pile To the hull side set and vertex set The minimum distance , the formula is: ， in, is the vertical distance from the center of the pile to the edge of the hull, is the Euclidean distance from the center of the pile to the vertex set of the hull; like Less than the safety threshold , triggering a collision warning and replanning the ship moving path based on the ant colony algorithm, updating the construction status of the pile in the digital twin model, and iteratively optimizing the construction sequence and path plan.

10. A system for dynamically optimizing the construction sequence of a water-sunk pile, for executing the method for dynamically optimizing the construction sequence of a water-sunk pile according to any one of claims 1 to 9, characterized in that: include: Data acquisition module, used to obtain four-dimensional dynamic multi-source data of the construction area in real time through multiple GPS positioning devices, geological detection equipment and environmental sensors; The construction sequence modeling module is used to build a mathematical model of the construction sequence based on pile position coordinates, geological condition parameters, and wharf structure parameters, and define a multi-objective optimization function for construction efficiency, cost, and safety risks; An intelligent algorithm optimization module uses a genetic algorithm to generate an initial plan for discrete sequence encoding of the pile construction sequence, an ant colony algorithm to model the ship movement path, and collaboratively optimizes the construction sequence and path plan based on a dynamic weight adjustment formula; A digital twin modeling module, which integrates ship parameters, designed pile parameters, and geographic data to construct a 3D dynamic twin model of the piling vessel, pile body, and construction environment; The collision detection module is used to calculate the minimum distance between the piling vessel and the constructed pile body through geometric algorithms, distinguish between square and round pile types, and output collision risk warnings; The dynamic planning module is used to trigger real-time adjustments to the construction sequence and ship movement path based on the collision detection results, update the digital twin model based on the optimized plan, and output dynamic construction instructions.

Images