High-performance wide flat river-sea direct ship form optimization method and system based on simulation driving design
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-03-17
Smart Images
Figure CN121683045A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of high-performance ship design, in particular to a high-performance wide and flat river-sea direct ship type optimization method and system based on simulation-driven design. BACKGROUND
[0002] River-sea direct ships are a type of ship developed to improve river-sea transportation efficiency by reducing intermediate links to reduce cargo losses and save time and transportation costs. River-sea direct ships not only can safely navigate in specific areas at sea, but also can adapt to river (river) channel depth, channel width and other navigation conditions, and are an important carrier for realizing the value of the Yangtze River golden waterway.
[0003] River-sea direct ships usually need to have special design and configuration to adapt to different water environments, and the selection of ship size and the type of ship body are also optimized according to the requirements of different water areas to ensure the navigability, safety and economy when navigating in rivers and oceans. Limited by the environmental conditions of the Yangtze River, the length and draft of the ship are limited, and the height is also restricted by the bridge clearance and difficult to break through. The only effective way to realize the demand of the shipping market is to increase the ship width and increase the width-depth ratio to develop a wide and flat ship type to increase the cargo capacity.
[0004] Wide and flat river-sea direct ships have very irregular ship type structural characteristics, and are subject to a wide range of wave environment changes. Under the action of high-frequency nonlinear wave loads, the bow structure is prone to slamming phenomenon, which can significantly increase the sailing resistance, reduce the speed and propulsion efficiency. On the other hand, river-sea direct ships need to adapt to the complex environment of narrow and shallow waterways in rivers and large waves in the sea, so the designed ship type needs to have good river section maneuverability and sea section seakeeping performance.
[0005] Currently, wide and flat river-sea direct ships exhibit good economy and adaptability in river-sea intermodal transport due to their unique ship type characteristics, but their wide and flat ship body, shallow draft, and large bow also bring challenges in the coordinated optimization of maneuverability, seakeeping performance, and speed performance. Traditional ship design methods usually rely on empirical formulas and local performance correction, and lack the ability to systematically simulate and multi-objective coordinated optimize the overall performance of wide and flat ship types in complex wave environments, especially in the coupling analysis of key issues such as maneuvering response, increased resistance in waves, motion stability, and slamming load.
[0006] Wide, flat hulls are prone to significant motion and wave impact in both headwinds and oblique waves. Furthermore, their flat bow structure and abrupt changes in cross-section can trigger severe localized impact loads and fluid-structure interaction problems upon impact. Existing prediction methods are mostly based on ideal fluid theory or simplified two-dimensional models, which struggle to accurately simulate the three-dimensional flow characteristics, bow air cushion effect, fluid jetting behavior, and their impact on maneuverability and seakeeping of wide, flat hulls in waves. Especially in the complex wave conditions of river-sea confluence areas, the interaction between the hull and waves is even more pronounced, further increasing the uncertainty in motion performance and load prediction, significantly impacting the ship's economic efficiency during navigation. Summary of the Invention
[0007] The purpose of this application is to provide a method and system for optimizing the hull form of high-performance wide and flat river-sea direct vessels based on simulation-driven design. This method can cover the entire process from hull form generation to performance optimization, thereby supporting iterative optimization and design decisions for high-performance wide and flat river-sea direct vessels.
[0008] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a high-performance, wide-bodied, flat-hulled river-sea direct vessel hull optimization method based on simulation-driven design, including: Obtain the characteristic parameters of the parent ship; the characteristic parameters include: bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concave angle; Based on the aforementioned feature parameters, geometric modeling is performed to obtain multiple candidate ship type schemes; Resistance prediction and slamming load simulation were performed on various candidate ship types to evaluate ship performance and obtain evaluation results. Based on the evaluation results, the improved PDPSO algorithm is used to optimize the scheme, resulting in the optimized ship type scheme.
[0009] In one embodiment, it further includes: The optimized hull design was tested and verified for its anti-smash performance.
[0010] In one embodiment, geometric modeling is performed based on the feature parameters to obtain multiple candidate ship type schemes, specifically including: The initial hull surface is generated using CAD software based on the aforementioned feature parameters; A smooth hull surface is constructed based on the initial hull surface using NURBS curves to meet the geometric compliance requirements; Interactive feature parameter adjustments were made to the smooth hull surface to reconstruct the hull surface, and trapezoidal profiles were used to optimize the wetted surface area distribution in order to reduce the proportion of wave-making resistance, resulting in a variety of candidate hull form schemes.
[0011] In one embodiment, the bow waterline inlet angle ranges from 15° to 30°; the stern fin tilt angle ranges from 5° to 15°; the tail fin width ranges from 0.8B to 1.2B; and the stern forward concave angle ranges from 8° to 12°.
[0012] In one embodiment, resistance prediction and slamming load simulation are performed on various candidate ship types to evaluate ship performance and obtain evaluation results, specifically including: Arbitrary Lagrange-Euler fluid-structure interaction algorithm was used to simulate slamming loads on various candidate ship types to determine the pressure distribution characteristics of the bow flare area and the bottom air cushion area; Resistance predictions are performed on various candidate ship types based on the Kriging surrogate model to obtain resistance prediction results. The resistance prediction results are used to characterize the trend of resistance performance changes within the design space. The Kriging surrogate model is constructed using fluid dynamics based on the sample ship types and their corresponding total resistance. The total resistance is determined based on residual resistance and frictional resistance. Based on the resistance prediction results and the pressure distribution characteristics information, the ship's performance is evaluated to obtain the evaluation results.
[0013] In one embodiment, the method for determining the Kriging proxy model specifically includes: Obtain sample information; the sample information includes: sample ship type design and corresponding total resistance; Construct the initial Kriging agent model; Based on the initial Kriging proxy model: For the sample ship type, the Reynolds-averaged Navier-Stokes equations method was adopted, combined with the VOF model to capture free surface deformation. StarCCM+ software was used for ship hydrodynamic simulation, and the k-ωSST model was selected to simulate near-wall flow and separated flow for hull flow calculation. Furthermore, based on preset convergence residuals, steady-state simulation was used to calculate resistance, yielding the resistance calculation results. The computational domain for the hull flow calculation must consider the properties of a wide, flat hull: flow direction range of -2.0 Lpp to +1.0 Lpp. The lateral range is ±2.0Lpp; the vertical range is -2.0Lpp to +1.0Lpp; Lpp is the length between perpendiculars; the origin is the intersection of the stern perpendicular and the waterline; the boundary conditions are set as follows: the inlet is a velocity inlet, the outlet is a pressure outlet, the hull surface is a no-slip wall, and the top and sides are symmetrical boundaries; the mesh strategy adopts a hybrid mesh technique, in which prismatic layer meshes are used around the hull to capture the boundary layer flow, and tetrahedral meshes are used in other locations; the drag calculation results are used to extract the total drag coefficient to decompose it into frictional drag and residual drag; Based on the resistance calculation results and the total resistance corresponding to the sample ship type scheme, the initial Kriging surrogate model is trained to obtain the Kriging surrogate model.
[0014] In one embodiment, based on the evaluation results, an improved PDPSO algorithm is used to optimize the scheme, resulting in an optimized ship type scheme, specifically including: Determine the initial parameters; the initial parameters include: setting the number of iterations and the dimension of the variables; Based on the aforementioned variable dimensions, an initial population is generated using Latin hypercube sampling; the initial population contains a set number of particles; the particles are candidate ship type schemes. Based on the parameter combination at the current iteration number, the velocity and position of each particle in the initial population are initialized using the Sobol sequence, and the inertial weight and learning factor at the current iteration number are determined; the parameter combination is the feature parameter. Based on the inertia weight and learning factor at the current iteration number, the real-time information of each particle at the current iteration number is determined; the real-time information includes update velocity and update position; the expression corresponding to the update velocity is: ; The expression corresponding to the update position is: ; in, For the first The first particle The speed of each iteration; For the first The first particle Dynamic inertia weights for each iteration; For the first The first particle The speed of each iteration; As the first learning factor; and All are random numbers in the range [0,1]. For the first The particle in the first The historical best position in the next iteration; As the second learning factor; For the first The globally optimal position found by the entire population in the next iteration; For the first The position of the next iteration; for Introducing the population diversity function of particle swarm optimization. ,when hour, ;when hour, ; For the first The population diversity function value of the next iteration; For the first The population diversity function value of the next iteration; For the first The first particle Dynamic inertia weights for each iteration; Based on population diversity Design nonlinear functions; ; ; The total number of particles; For the first The second iteration The average Euclidean distance between each particle and other particles; For all The average value; For the first The population diversity function value of the next iteration; Based on the evaluation results, a particle swarm fitness function is determined to calculate the comprehensive performance index value of each particle. Based on the real-time information and the comprehensive performance index value, the population diversity information at the current iteration number is determined; wherein, crowding calculation and adaptive mutation mechanism are introduced, and when the population diversity information at the current iteration number is lower than a preset threshold, mutation operation is performed on some of the particles; the population diversity information includes: individual extreme values and global extreme values; Determine whether the iteration stopping condition is met; the iteration stopping condition is that the current iteration count has reached a set iteration count. If the iteration stopping condition is met, the parameter combination at the current iteration number is taken as the optimal parameter combination; Based on the optimal parameter combination and the corresponding population diversity information, the optimized ship type scheme is determined. If the iteration stopping condition is not met, the parameter combination for the current iteration number is updated based on the comprehensive performance index value for the current iteration number, and the updated parameter combination is used as the parameter combination for the next iteration number. The process then returns to the step of "initializing the velocity and position of each particle in the initial population based on the Sobol sequence according to the parameter combination for the current iteration number, and determining the inertia weight and learning factor for the current iteration number".
[0015] In one embodiment, the expression corresponding to the particle swarm fitness function is: ; in, This refers to the comprehensive performance index value; This is the total drag coefficient; This refers to the deviation rate of the slewing diameter. The effective wave height of the heave; , and All are weighting coefficients.
[0016] In one embodiment, the optimized hull design undergoes slamming resistance testing and verification, specifically including: The optimized ship design was modeled at a scale of 1:25 to create a wooden model; the wooden model was made of birch laminate. The center of gravity of the wooden model was adjusted by counterweight and the surface of the hull was smoothed. The processed model is fixed to the test tower, and the water entry angle is adjusted to the preset value through the hydraulic system. The release height is set according to the working conditions. The experiment involves data acquisition and analysis of experimental results. Data acquisition includes recording the water entry process of the processed model using a high-speed camera, measuring the impact pressure using a dynamic pressure sensor, and measuring the motion response using an acceleration sensor. The analysis of experimental results includes statistical analysis of the peak impact pressure, analysis of pressure distribution characteristics, and evaluation of the air cushion effect.
[0017] Secondly, this application provides a high-performance, wide-bodied, flat-hulled river-sea direct-access vessel hull optimization system based on simulation-driven design, comprising: The feature parameter acquisition module is used to acquire the feature parameters of the parent ship; the feature parameters include: bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concave angle; The modeling processing module is used to perform geometric modeling processing based on the feature parameters to obtain multiple candidate ship type schemes; The performance evaluation module is used to predict resistance and simulate slamming loads for various candidate ship types to evaluate ship performance and obtain evaluation results. The optimization module is used to optimize the scheme based on the evaluation results using the improved PDPSO algorithm to obtain the optimized ship type scheme.
[0018] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application provides a high-performance, wide-bodied, flat-hulled river-sea direct-access vessel hull optimization method and system based on simulation-driven design. It integrates parametric geometric modeling, simulation, and an improved particle swarm optimization algorithm to establish a multi-objective collaborative hull optimization design method for wide-bodied, flat-hulled river-sea direct-access vessels. By parametrically controlling key parameters of the bow and stern lines, characteristic parameters of the parent vessel are obtained. These characteristic parameters include: bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concavity angle. Combined with resistance prediction and slamming load simulation, and driven by an improved Population Diversity-based Particle Swarm Optimization (PDPSO) algorithm, the optimized hull hull scheme is obtained by iteratively optimizing design variables (i.e., characteristic parameters). This achieves full-process coverage from hull generation to performance optimization, supporting iterative optimization and design decisions for high-performance, wide-bodied, flat-hulled river-sea direct-access vessels. Attached Figure Description
[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0020] Figure 1 A flowchart of a high-performance, wide-bodied, flat-hulled direct-ship vessel hull optimization method based on simulation-driven design; Figure 2 For parametric design flowchart; Figure 3 Design a schematic diagram for the variables corresponding to the features; Figure 4 A schematic diagram of the improved PDPSO algorithm based on Sobol sequences; Figure 5 This is a structural diagram of a high-performance, wide-short-faced, direct-ship river-sea vessel hull optimization system based on simulation-driven design. Detailed Implementation
[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0022] To improve the overall design performance of wide-slender river-sea direct vessels in terms of maneuverability, seakeeping, and speed, this application proposes a hull design method based on multidisciplinary simulation and performance coupling analysis, addressing a key technical challenge for those skilled in the art. This application presents a high-performance wide-slender river-sea direct vessel hull design based on Simulation Based Design (SBD), integrating parametric geometric modeling, computational fluid dynamics (CFD) and potential-flow coupling calculations, motion response prediction, and structural load analysis. This achieves full-process coverage from hull generation to multi-objective performance optimization, thereby supporting rapid iteration and design decisions for high-performance wide-slender river-sea direct vessels.
[0023] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0024] In one exemplary embodiment, such as Figure 1 As shown, a high-performance, wide-bodied, flat-hulled river-sea direct vessel hull optimization method based on simulation-driven design is provided, including: Step 100: Obtain the characteristic parameters of the parent ship. The characteristic parameters include: bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concave angle.
[0025] The bow waterline inlet angle ranges from 15° to 30°; the stern fin inclination angle ranges from 5° to 15°; the tail fin width ranges from 0.8B to 1.2B; and the stern forward concave angle ranges from 8° to 12°. B typically refers to the beam, which is the maximum horizontal distance between the rib lines at midships (or the maximum transverse width at the design waterline).
[0026] Step 200: Perform geometric modeling based on feature parameters to obtain multiple candidate ship type schemes.
[0027] Among these, geometric modeling based on feature parameters yields a variety of candidate ship type schemes, including: The initial hull surface was generated using CAD software based on feature parameters. A smooth hull surface was constructed based on the initial hull surface using NURBS curves to meet the requirements of geometric compliance. The smooth hull surface was interactively adjusted for feature parameters to reconstruct the hull surface. A trapezoidal profile was used to optimize the wetted surface area distribution to reduce the proportion of wave-making resistance, resulting in a variety of candidate hull form schemes.
[0028] Step 300: Perform resistance prediction and slamming load simulation on multiple candidate ship types to evaluate ship performance and obtain evaluation results.
[0029] This includes resistance prediction and slamming load simulation for various candidate ship designs to evaluate ship performance and obtain evaluation results, specifically including: Arbitrary Lagrange-Euler fluid-structure interaction algorithm was used to simulate slamming loads on various candidate ship types to determine the pressure distribution characteristics of the bow overhang region and the bottom air cushion region.
[0030] Resistance predictions were performed on various candidate ship types based on the Kriging surrogate model, and the resistance prediction results were obtained. The resistance prediction results are used to characterize the trend of resistance performance changes within the design space. The Kriging surrogate model is constructed using fluid dynamics based on the sample ship type and the corresponding total resistance. The total resistance is determined based on the residual resistance and frictional resistance.
[0031] Ship performance is evaluated based on resistance prediction results and pressure distribution characteristics, and the evaluation results are obtained.
[0032] As an optional implementation method, the method for determining the Kriging proxy model specifically includes: Obtain sample information; sample information includes: sample ship type scheme and corresponding total resistance.
[0033] Construct the initial Kriging proxy model; based on the initial Kriging proxy model: For the sample ship type, the Reynolds-averaged Navier-Stokes equations method was adopted, combined with the VOF model to capture free surface deformation. StarCCM+ software was used for ship hydrodynamic simulation, and the k-ωSST model was selected to simulate near-wall flow and separated flow for hull flow calculation. Furthermore, based on preset convergence residuals, steady-state simulation was used to calculate resistance, yielding the resistance calculation results. The computational domain for the hull flow calculation must consider the properties of a wide, flat hull: the flow direction range is -2.0Lpp to +1.0Lp. p; Lateral range ±2.0Lpp; Vertical range -2.0Lpp to +1.0Lpp; Lpp is the length between perpendiculars; Origin is the intersection of the stern perpendicular and the waterline; Boundary conditions are set as follows: inlet is a velocity inlet, outlet is a pressure outlet, hull surface is a no-slip wall, and top and sides are symmetrical boundaries; The mesh strategy adopts a hybrid mesh technique, in which prismatic layer meshes are used around the hull to capture boundary layer flow, and tetrahedral meshes are used at other locations; The drag calculation results are used to extract the total drag coefficient to decompose it into frictional drag and residual drag.
[0034] Based on the resistance calculation results and the total resistance corresponding to the sample ship type scheme, the initial Kriging surrogate model is trained to obtain the Kriging surrogate model.
[0035] Step 400: Based on the evaluation results, the improved PDPSO algorithm is used to optimize the scheme and obtain the optimized ship type scheme.
[0036] Based on the evaluation results, the improved PDPSO algorithm is used to optimize the scheme, resulting in the optimized ship type scheme, which specifically includes: Determine the initial parameters; the initial parameters include: setting the number of iterations and the dimensions of the variables.
[0037] Based on the variable dimension, an initial population is generated using Latin hypercube sampling; the initial population contains a set number of particles; the particles are candidate ship type schemes.
[0038] Based on the parameter combination at the current iteration number, the velocity and position of each particle in the initial population are initialized using the Sobol sequence, and the inertial weight and learning factor at the current iteration number are determined; the parameter combination is the feature parameter.
[0039] Based on the inertia weight and learning factor at the current iteration number, the real-time information of each particle at the current iteration number is determined; the real-time information includes update velocity and update position; the expression corresponding to the update velocity is: ; The expression corresponding to the update position is: ; in, For the first The first particle The speed of each iteration; For the first The first particle Dynamic inertia weights for each iteration; For the first The first particle The speed of each iteration; As the first learning factor; and All are random numbers in the range [0,1]. For the first The particle in the first The historical best position in the next iteration; As the second learning factor; For the first The globally optimal position found by the entire population in the next iteration; For the first The position of the next iteration; for Introducing the population diversity function of particle swarm optimization. ,when hour, ;when hour, ; For the first The population diversity function value of the next iteration; For the first The population diversity function value of the next iteration; For the first The first particle Dynamic inertia weights for each iteration; Based on population diversity Design nonlinear functions; ; ; The total number of particles; For the first The second iteration The average Euclidean distance between each particle and other particles; For all The average value; For the first The population diversity function value of the next iteration.
[0040] The particle swarm fitness function is determined based on the evaluation results to calculate the comprehensive performance index value of each particle.
[0041] The expression for the particle swarm fitness function is: .
[0042] in, This refers to the comprehensive performance index value; This is the total drag coefficient; This refers to the deviation rate of the slewing diameter. The effective wave height of the heave; , and All are weighting coefficients.
[0043] Based on real-time information and comprehensive performance index values, the population diversity information at the current iteration number is determined. Among them, crowding degree calculation and adaptive mutation mechanism are introduced. When the population diversity information at the current iteration number is lower than a preset threshold, some particles are mutated. The population diversity information includes individual extreme values and global extreme values.
[0044] Determine if the iteration stopping condition is met; the iteration stopping condition is that the current iteration count has reached the set iteration count.
[0045] If the iteration stopping condition is met, the parameter combination at the current iteration number is taken as the optimal parameter combination. Based on the optimal parameter combination and the corresponding population diversity information, the optimized ship type scheme is determined.
[0046] If the iteration stopping condition is not met, the parameter combination for the current iteration number is updated based on the comprehensive performance index value for the current iteration number, and the updated parameter combination is used as the parameter combination for the next iteration number. The process then returns to the step of "initializing the velocity and position of each particle in the initial population based on the Sobol sequence according to the parameter combination for the current iteration number, and determining the inertial weight and learning factor for the current iteration number".
[0047] As an optional implementation, the method further includes: conducting experimental verification of the slamming resistance performance of the optimized ship design.
[0048] The optimized hull design underwent slamming resistance testing and verification, specifically including: The optimized ship design was modeled at a scale of 1:25 to create a wooden model; the material for the wooden model was birch laminate.
[0049] The center of gravity of the wooden model was adjusted by counterweight and the surface of the hull was smoothed. The treated model was fixed to the test tower, and the water entry angle was adjusted to the preset value through the hydraulic system. The release height was set according to the working conditions.
[0050] The experiment involved data acquisition and analysis of the results. Data acquisition included recording the water entry process of the processed model using a high-speed camera, measuring the impact pressure using a dynamic pressure sensor, and measuring the motion response using an accelerometer. The results analysis included statistical analysis of the peak impact pressure, analysis of pressure distribution characteristics, and evaluation of the air cushion effect.
[0051] The main principle of this application is based on a simulation-driven design framework, integrating parametric geometric modeling, computational fluid dynamics, an improved particle swarm optimization algorithm based on population diversity, and model testing verification technology. For wide, flat, river-sea direct-access vessels, it establishes a multi-objective collaborative hull optimization design method encompassing maneuverability, seakeeping, and speed. By parametrically controlling key bow and stern line parameters, combined with CFD resistance prediction and slamming load simulation, the optimization algorithm iterates through design variables. Finally, slamming resistance is verified through scaled-down model free-fall tests, thereby improving the overall sailing performance of wide, flat vessels. The specific technical solution is as follows: The overall process of this application includes four main stages: parametric modeling, CFD simulation analysis, multi-objective optimization, and experimental verification. First, by parametrically adjusting key parameters of the bow and stern hull forms (such as the bow waterline inflow angle and stern fin inclination angle), multiple candidate hull forms are generated. Then, CFD is used for resistance prediction and slamming load simulation to evaluate the performance of each hull form. Next, an improved PDPSO algorithm is used for multi-objective optimization to find the optimal parameter combination. Finally, slamming resistance is verified through free-fall tests on scaled-down models, thereby improving the overall sailing performance of the wide and flat hull form.
[0052] (1) The parametric modeling process adopts a feature-based design method, using NURBS (Non-Uniform Rational B-Spline) curves as the core geometric expression tool. The parametric design process is as follows: Figure 2 As shown, the specific process includes: (1) extracting characteristic parameters and model value points from the parent ship; (2) generating bow and stern characteristic curves using NURBS curves; (3) reconstructing the hull surface through interactive parameter adjustment; and (4) optimizing the wetted surface area distribution using trapezoidal profiles to reduce the proportion of wave-making resistance.
[0053] The characteristic parameters were determined taking into account the special performance requirements of wide and flat river-sea direct vessels, such as... Figure 3 As shown in Table 1. The design variable, i.e. the characteristic parameter, is the bow waterline inlet angle. This primarily affects a ship's seakeeping ability. By controlling this angle within the range of 15°-30°, deck swell can be effectively reduced; stern fin tilt angle The angle affects the wake field structure; an excessively large angle will intensify flow separation, while an excessively small angle will hinder the effectiveness of the rudder; tail fin width The proportional relationship with the beam B directly affects the ship's maneuverability; the stern forward concave angle This mainly affects the flow separation characteristics at the stern, which is of great significance for reducing resistance.
[0054] Table 1: Range of Design Variables and Core Parameters
[0055] (2) The numerical simulation adopted the Reynolds-averaged Navier-Stokes equations (RANS) method, combined with the VOF (Volume of Fluid) model to capture free surface deformation. StarCCM+ was selected as the software tool, which has good accuracy and reliability in ship hydrodynamic simulation. The k-ω SST (Shear Stress Transport) model was selected as the turbulence model, which can accurately simulate near-wall flow and separated flow, and is suitable for flow calculations around the hull.
[0056] The computational domain was designed to fully consider the unique characteristics of a wide, flat hull: the flow (x) range was -2.0Lpp to +1.0Lpp (Lpp being the length between perpendiculars), the lateral (y) range was ±2.0Lpp, and the vertical (z) range was -2.0Lpp to +1.0Lpp (the origin being the intersection of the stern perpendicular and the waterline). Boundary conditions were set as follows: the inlet was a velocity inlet, the outlet a pressure outlet, the hull surface was a no-slip wall, and the top and sides were symmetrical boundaries. A hybrid meshing strategy was employed, using a prismatic layer mesh around the hull to capture the boundary layer flow, and a tetrahedral mesh at a distance to reduce computational load. The wall y+ was greater than 30, the wall growth rate was 1.05, and the volumetric mesh count was approximately 5 million. This meshing configuration ensured computational accuracy while controlling computational costs.
[0057] Resistance performance prediction involved generating 20 sample ship types using a uniform experimental design to calculate the total resistance (residual resistance + frictional resistance). Based on the calculation results, a Kriging surrogate model was established, which effectively predicts the trend of resistance performance changes within the design space, providing guidance for the optimization algorithm. Slamming load simulation employed an arbitrary Lagrange-Euler (ALE) fluid-structure interaction algorithm to analyze the pressure distribution characteristics of the bow flare region and the bottom air cushion region.
[0058] (3) The PDPSO algorithm based on Sobol sequences is used for scheme optimization. The speed update and position update formulas of the algorithm are as follows: .
[0059] .
[0060] Compared to the standard particle swarm optimization algorithm, three improvements are made to obtain the optimal parameters for ship hull optimization: (a) Dynamic inertia weight: linearly decreases from 0.9 to 0.4 with the number of iterations, balancing global exploration and local exploitation capabilities; (b) Adaptive learning factor: adjusts social cognition (i.e., based on population diversity) according to population diversity. ) and individual cognition (i.e. (c) Weighting to avoid premature convergence; (d) Constraint handling: Introduce a penalty function to handle constraints such as bow outward angle (≤25°) to control the risk of slamming.
[0061] The objective function is to minimize the overall performance index: .
[0062] This is the total drag coefficient; This refers to the deviation rate of the slewing diameter. The effective wave height of the heave; Weighting coefficients ( , , ).
[0063] The weighting coefficients reflect the degree of emphasis placed on various performance indicators by the wide and flat river-sea direct vessel type. Among them, speed (resistance performance) is the most important, followed by maneuverability and seakeeping.
[0064] The optimization algorithm uses Latin hypercube sampling to generate the initial population, ensuring uniform exploration of the design space. The initial number of particles and the number of iterations are set to control computational cost while maintaining optimization effectiveness. In each iteration, the algorithm obtains fitness values by querying the surrogate model, updates individual and global extrema (i.e., updates the individual and swarm optimal solutions), and finally outputs the optimal parameter combination (i.e., the optimal solution). The algorithm flow is as follows: Figure 4 As shown.
[0065] (4) Conduct anti-slamming performance tests to verify the model design. The model design criteria are based on similarity theory to ensure geometric, kinematic, and dynamic similarity between the model and the actual ship. The scaling ratio λ = 1:25 satisfies Froude number (Fr) similarity (Fr = 0.2~0.4) and Reynolds number (Re) similarity (Re = 1.2 × 10⁻⁶). 6 The model is made of wood, and the error in the center of gravity position is controlled within ≤2% to ensure the reliability of the test results.
[0066] The sensor arrangement includes: (1) Pressure sensor: 8 measuring points are arranged in the outward flare area of the bow (1.5m from the bowpost) and 4 measuring points are arranged in the bottom air cushion area to measure the spatiotemporal distribution of the slamming pressure; (2) Acceleration sensor: a three-dimensional acceleration sensor is arranged at the center of gravity to measure the ship's motion response; (3) Photoelectric switch and electronic digital display angle meter: measure the entry speed and angle, with an accuracy of ±0.1°.
[0067] The test conditions were designed to account for slamming characteristics under different navigation conditions. As shown in Table 2, slamming phenomena at low, medium, and high speeds were simulated by varying the drop height and entry angle. Each test was repeated five times to ensure the statistical reliability of the data.
[0068] Table 2: Range of Design Variables and Core Parameters
[0069] This application successfully solves the multi-objective performance balancing challenge of wide, flat river-sea direct-access vessels by employing an integrated parametric design-simulation-experiment approach. Compared to traditional design methods, this application resolves the contradictions between speed, maneuverability, and seakeeping in wide, flat vessel types through multidisciplinary coupled analysis, particularly addressing the unique slamming flutter and wave-induced vibration issues of river-sea direct-access vessels. This application provides a complete performance optimization design process and quantitative design tools for wide, flat river-sea direct-access vessels, effectively improving their economic and social benefits.
[0070] First, a parametric geometric model of the hull is established. The first step is to define the bow and stern line parameters in CAD software (such as CATIA or Rhino) to generate the initial hull surfaces. Key parameters include the bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concavity angle. Figure 3 The second step is to construct a smooth hull surface using NURBS curves to ensure geometric compliance. The third step is to generate 20 sets of variant hull designs through interactive parameter adjustment. The parameter adjustment range is determined based on orthogonal experimental design to ensure that the samples cover the design space and are representative. After each parameter adjustment, an STL format geometry file is exported for subsequent CFD analysis. The entire process is automated using Python scripts to improve modeling efficiency.
[0071] Then, CFD simulation calculations were performed, including impact resistance and drag performance. The impact resistance numerical simulation used StarCCM+ to establish a three-dimensional finite element model of the wide, flat bow structure, fluid Eulerian elements for air and water, and boundary conditions. Considering both computational efficiency and accuracy, a Lagrange model mesh size of 2mm (approximately 0.0032 times the hull profile width) was adopted. A uniform mesh (4mm) was used within the main impact Eulerian domain, while a gradient mesh was used in the more distant Eulerian domains. Symmetrical boundary conditions were established at the Y=0 plane to significantly reduce the number of meshes. Non-reflective boundary conditions were used at the fluid bottom and boundaries to eliminate the influence of stress wave reflection. The ALE fluid-structure interaction algorithm was used to analyze the three-dimensional effect of the bow entering the water and impacting the bow. The drag performance numerical simulation settings involved importing the STL file into StarCCM+ for mesh generation and calculation settings. A cut-body mesh was used with a base size of 0.05Lpp. The mesh around the hull was refined, with 5 prism layers, resulting in a total mesh count of approximately 5 million. The time step was set to 0.01 s to satisfy the CFL condition. Boundary conditions were set as follows: the inlet was a velocity inlet, with the incoming flow velocity set according to the Froude number (Fr=0.28); the outlet was a pressure outlet; the symmetry plane was set as a symmetric boundary condition; and the hull surface was a no-slip wall. The k-ω SST model was selected as the turbulence model, which can accurately simulate the flow separation phenomenon around the hull. Drag calculations were performed through steady-state simulation, with the convergence residual set to ≤1e. -4 After each calculation, the total drag coefficient C is extracted. t The resistance is decomposed into frictional resistance and residual resistance components. Based on the calculation results of 20 sample ship types, a Kriging surrogate model is established, which can accurately predict the resistance performance at any point in the design space.
[0072] Further parameter optimization iterations are performed. The initial population of the optimization algorithm is generated using Latin hypercube sampling to ensure uniform coverage of the design space. Initial settings include a particle count of 30, 50 iterations, and a design variable dimension of 4 (corresponding to the four key parameters). The inertia weight has an initial value of 0.9 and a final value of 0.4; the first learning factor... Second learning factor The initial value is 2.0, adaptively adjusted based on population diversity. Fitness evaluation is performed using a surrogate model, calculating the comprehensive performance index F for each particle. After each iteration, the individual best (pbest) and global best (gbest) are updated. To maintain population diversity, crowding calculation and an adaptive mutation mechanism are introduced; when population diversity falls below a threshold, some particles are mutated. The optimization process converges after 50 iterations, outputting the optimal parameter combination. For a wide, flat-hulled river-sea direct vessel, the optimized combination yields a bow waterline inflow angle of 22°, a stern fin inclination angle of 10°, a tail fin width of 1.05B, and a stern forward concave angle of 10°. This parameter combination balances speed, maneuverability, and seakeeping requirements, achieving the best overall performance.
[0073] Finally, model tests were conducted for verification. The model was fabricated using a wooden model at a scale of λ=1:25, with birch laminate as the material to ensure structural strength and rigidity. The model's center of gravity was adjusted using counterweights, with the error controlled to ≤2%. The hull surface was smoothed to reduce the impact of roughness on the test results. The test setup involved fixing the model to a dedicated test tower and adjusting the water entry angle to preset values (5°, 10°, 15°) using a hydraulic system. The release height was set according to the working conditions (0.5m, 1.0m, 1.5m). The data acquisition system included a high-speed camera (1000fps) recording the water entry process, a dynamic pressure sensor measuring the impact pressure, and an acceleration sensor measuring the motion response. The test results analysis included statistical analysis of the peak impact pressure, pressure distribution characteristics analysis, and air cushion effect assessment. The maximum pressure value in the outward-flaring area of the bow model measured in the test was ≤45kPa, meeting the structural strength requirements. The pressure distribution uniformity in the bottom air cushion area was improved by 20% compared to before optimization, indicating that the optimized design can effectively improve the air cushion effect and reduce the impact of slam loads on the hull structure.
[0074] In one exemplary embodiment, such as Figure 5 As shown, a high-performance, wide-bodied, flat-hulled river-sea direct vessel hull optimization system based on simulation-driven design is provided, including: The feature parameter acquisition module is used to acquire the feature parameters of the parent ship; the feature parameters include: bow waterline inflow angle, stern fin inclination angle, tail fin width, and stern forward concave angle.
[0075] The modeling processing module is used to perform geometric modeling processing based on the feature parameters to obtain multiple candidate ship type schemes.
[0076] The performance evaluation module is used to predict resistance and simulate slamming loads for various candidate ship types to evaluate ship performance and obtain evaluation results.
[0077] The optimization module is used to optimize the scheme based on the evaluation results using the improved PDPSO algorithm to obtain the optimized ship type scheme.
[0078] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0079] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design, characterized in that, The method comprises the following steps: obtaining characteristic parameters of a mother ship; the characteristic parameters include a bow waterline inflow angle, a stern fin inclination angle, a tail fin width and a ship stern fore recess angle; performing geometric modeling processing based on the characteristic parameters to obtain multiple candidate ship type schemes; performing resistance prediction and slamming load simulation on the multiple candidate ship type schemes to perform ship type performance evaluation and obtain evaluation results; based on the evaluation results, using an improved PDPSO algorithm to perform scheme optimization to obtain an optimized ship type scheme.
2. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 1, characterized in that, Further comprising: performing test verification processing on the optimized ship type scheme to verify the anti-slugging performance.
3. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 1, characterized in that, The geometric modeling processing based on the characteristic parameters to obtain multiple candidate ship type schemes specifically comprises: generating an initial ship body surface based on the characteristic parameters using CAD software; constructing a smooth ship body surface based on the initial ship body surface using NURBS curve to meet the geometric fairing requirement; reconstructing the smooth ship body surface by adjusting the interactive characteristic parameters, and optimizing the wet surface area distribution by using trapezoidal section to reduce the wave-making resistance ratio to obtain multiple candidate ship type schemes.
4. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 1, characterized in that, The bow waterline inflow angle ranges from 15° to 30°, the stern fin inclination angle ranges from 5° to 15°, the tail fin width ranges from 0.8B to 1.2B, and the ship stern fore recess angle ranges from 8° to 12°.
5. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 1, characterized in that, The resistance prediction and slamming load simulation on the multiple candidate ship type schemes to perform ship type performance evaluation and obtain evaluation results specifically comprises: performing slamming load simulation on the multiple candidate ship type schemes using arbitrary Lagrangian-Eulerian fluid-structure coupling algorithm to determine the pressure distribution characteristic information of the bow outflow area and the bottom air cushion area; performing resistance prediction on the multiple candidate ship type schemes based on the Kriging surrogate model to obtain resistance prediction results; the resistance prediction results are used to represent the resistance performance variation trend in the design space; the Kriging surrogate model is constructed based on the sample ship type scheme and the corresponding total resistance using fluid mechanics; the total resistance is determined according to the residual resistance and the friction resistance; performing ship type performance evaluation based on the resistance prediction results and the pressure distribution characteristic information to obtain evaluation results.
6. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 5, characterized in that, The determination method of the Kriging surrogate model specifically comprises: obtaining sample information; the sample information includes sample ship type schemes and corresponding total resistance; constructing an initial Kriging surrogate model; based on the initial Kriging surrogate model: For the sample ship type scheme, a Reynolds averaged Navier-Stokes equation method is adopted, a VOF model is combined to capture the free surface deformation, StarCCM+ software is used for ship hydrodynamics simulation, a k-omega SST model is selected to simulate the near-wall region flow and separation flow to perform ship hull flow calculation, and based on a preset convergence residual, a steady-state simulation is performed to calculate the resistance, and a resistance calculation result is obtained; wherein the calculation domain for the ship hull flow calculation needs to consider the properties of the wide and flat ship type: the streamwise range is -2.0Lpp to +1.0Lpp; the lateral range is ±2.0Lpp; the vertical range is -2.0Lpp to +1.0Lpp; Lpp is the length between perpendiculars; the origin is the intersection point of the stern perpendicular and the waterline; the boundary conditions are set as: the inlet is a velocity inlet, the outlet is a pressure outlet, the ship hull surface is a no-slip wall, and the top and sides are symmetric boundaries; the mesh strategy adopts a hybrid mesh technology, wherein the prismatic layer mesh is used around the ship hull to capture the boundary layer flow, and the tetrahedral mesh is used at the remaining positions; the resistance calculation result is used to extract the total resistance coefficient to decompose into friction resistance and residual resistance; According to the total resistance corresponding to the sample ship type scheme, the initial Kriging surrogate model is trained to obtain the Kriging surrogate model.
7. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 1, characterized in that, Based on the evaluation result, an improved PDPSO algorithm is used for scheme optimization to obtain an optimized ship type scheme, specifically including: determining initial parameters; the initial parameters include: setting the number of iterations and the variable dimension; based on the variable dimension, using Latin hypercube sampling to generate an initial population; the initial population contains a set number of particles; the particles are candidate ship type schemes; according to the parameter combination under the current number of iterations, initializing the speed and position of each particle in the initial population based on the Sobol sequence, and determining the inertia weight under the current number of iterations and the learning factor under the current number of iterations; the parameter combination is the characteristic parameter; determining the real-time information of each particle under the current number of iterations according to the inertia weight under the current number of iterations and the learning factor under the current number of iterations; the real-time information includes updated speed and updated position; the expression corresponding to the updated speed is: ； the expression corresponding to the updated position is: ； in, For the first The first particle The speed of each iteration; For the first The first particle Dynamic inertia weights for each iteration; For the first The first particle The speed of each iteration; As the first learning factor; and All are random numbers in the range [0,1]. For the first The particle in the first The historical best position in the next iteration; As the second learning factor; For the first The globally optimal position found by the entire population in the next iteration; For the first The position of the next iteration; For , a population diversity function of the particle swarm is introduced when , ; when , ; the population diversity function value for the first iteration; the population diversity function value for the first iteration; the population diversity function value for the first iteration; the population diversity function value for the first iteration; the dynamic inertia weight for the first iteration of the first particle; the dynamic inertia weight for the first iteration of the first particle; the dynamic inertia weight for the first iteration of the first particle; designing a non-linear function based on population diversity designing a non-linear function based on population diversity ； ； The total number of particles; For the first The second iteration The average Euclidean distance between each particle and other particles; For all The average value; For the first The population diversity function value of the next iteration; determining the particle swarm fitness function based on the evaluation result to calculate the comprehensive performance index value of each particle; based on the real-time information, determining the population diversity information under the current number of iterations based on the comprehensive performance index value; wherein a crowding degree calculation and an adaptive mutation mechanism are introduced, and when the population diversity information under the current number of iterations is lower than a preset threshold, a mutation operation is performed on part of the particles; the population diversity information includes individual extreme value and global extreme value; determining whether the iteration stop condition is met; the iteration stop condition is that the current number of iterations reaches the set number of iterations; if the iteration stop condition is met, the parameter combination under the current number of iterations is taken as the optimal parameter combination; based on the optimal parameter combination and the corresponding population diversity information, an optimized ship type scheme is determined; If the iteration stop condition is not met, the parameter combination at the current iteration number is updated according to the comprehensive performance index value at the current iteration number, and the updated parameter combination is taken as the parameter combination at the next iteration number, and the step of initializing the speed and position of each particle in the initial population based on the Sobol sequence according to the parameter combination at the current iteration number, and determining the inertia weight at the current iteration number and the learning factor at the current iteration number is returned.
8. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 7, characterized in that, The expression corresponding to the particle swarm fitness function is: ； wherein, is a comprehensive performance index value; is a total resistance coefficient; is a deviation rate of the turning diameter; is an effective wave height of the heave; , and are all weight coefficients.
9. The high-performance wide and flat river-sea direct ship hull form optimization method based on simulation-driven design according to claim 2, characterized in that, The optimized ship type scheme is subjected to experimental verification processing of anti-pounding performance, specifically including: The optimized ship type scheme is subjected to model processing according to a scale ratio of 1:25 to manufacture a wooden model; the material of the wooden model is selected from birch laminated wood; The center of gravity of the wooden model is subjected to weight adjustment and the surface of the ship body is subjected to smoothing processing; The processed model is fixed to a test tower, the water entry angle is adjusted to a preset value by a hydraulic system, and the release height is set according to the working condition requirement; Experimental data acquisition and experimental result analysis are performed; the experimental data acquisition includes: recording the water entry process of the processed model based on a high-speed camera, measuring the slamming pressure based on a dynamic pressure sensor, and measuring the motion response based on an acceleration sensor; the experimental result analysis includes slamming pressure peak value statistics, pressure distribution characteristic analysis and air cushion effect evaluation.
10. A high-performance wide and flat river-sea direct ship hull form optimization system based on simulation-driven design, characterized in that, It includes: A characteristic parameter acquisition module is configured to acquire characteristic parameters of a mother ship; The characteristic parameters include: a bow waterline inflow angle, a stern fin inclination angle, a tail fin width and a ship stern front concave angle; A modeling processing module is configured to perform geometric modeling processing based on the characteristic parameters to obtain a plurality of candidate ship type schemes; A performance evaluation module is configured to perform resistance prediction and slamming load simulation on the plurality of candidate ship type schemes to perform ship type performance evaluation and obtain evaluation results; An optimization module is configured to perform scheme optimization based on the evaluation results by using an improved PDPSO algorithm to obtain an optimized ship type scheme.