High-efficiency and high-fidelity coupling simulation method and simulation system for floating wind turbine

Abstract

The invention discloses an efficient and high-fidelity coupling simulation method for a floating wind turbine. The method comprises the following steps: determining a current position of a hub reference point under global coordinates based on a rigid body posture; the unit where the hub position is located and the process of the hub position are determined in the parallel partition grids; in the process reading unit, the relative wind speed at the hub is constructed to be the linear speed obtained by subtracting the linear speed obtained by multiplying the translating speed and the angular speed of the rigid body by the vector diameter from the local flow speed; the reference axis is converted into the current axis direction of the fan according to the quaternion posture, and the axial component of the relative wind speed is obtained; performing fast binary retrieval and linear interpolation on the axial wind speed according to a preset wind speed-thrust discrete curve to obtain a thrust scalar and generate a thrust vector along the axis direction; the vector diameter of the hub relative to the rotation center and the thrust vector are subjected to cross multiplication to obtain torque; and the thrust and the torque are combined with the fluid integral force and the gravity to serve as external force and external torque of a 6DoF rigid body kinetic equation, and the thrust pose is obtained. According to the method, explicit modeling impeller geometry and body-fitted grid division thereof are avoided, the method is different from an actuating line model, the model has high robustness while the calculation cost is reduced, meanwhile, the gas-wave-mooring coupling precision is guaranteed, and the method is suitable for the engineering-scale parallel calculation environment.

Description

Technical Field

[0001] This invention relates to a high-efficiency, high-fidelity coupled simulation method and simulation system for floating wind turbines. Background Technology

[0002] Floating wind turbines generate complex coupled responses under the influence of wind, waves, and currents, involving the interaction of multiple physical fields such as aerodynamics, fluid dynamics, rigid body dynamics, and mooring dynamics. As offshore wind power develops towards deeper waters, the design, analysis, and optimization of floating wind turbines place higher demands on high-fidelity numerical simulation technology.

[0003] In existing technologies, full-scale simulation of blade and tower geometry using direct CFD analysis is the main method for obtaining high-precision results. This method discretizes the blades, tower, and surrounding flow field using fine meshes, which can accurately capture the aerodynamic characteristics of the impeller, wake effects, and coupling with the platform motion. There are some technical problems: (1) Huge computational cost: For a typical 5-15MW floating wind turbine, the blade length can reach 60-120 meters and the tower height can reach 80-150 meters. Full-scale CFD simulation usually requires tens of millions to hundreds of millions of grid cells, and the simulation time can reach several days to several weeks, resulting in huge computational resource consumption; (2) Limited parallel scalability: Due to the complex blade geometry and dense boundary layer mesh, the parallel efficiency usually drops significantly after 96-256 cores, making it difficult to fully utilize large-scale parallel computing resources; (3) Complex mesh generation: The complex geometry of the blade and tower makes mesh generation difficult, especially the quality control of the boundary layer mesh, which requires a lot of manual intervention; (4) Time step limitation: In order to capture blade rotation and vortex shedding, the time step usually needs to be limited to about 10^-3 seconds, which further increases the computational cost.

[0004] To reduce computational costs, existing technologies employ simplified methods such as the ActuatorDiskModel (ADM) or ActuatorLineModel (ALM) to simulate the aerodynamic effects of the impeller by introducing distributed forces into the flow field. However, this method also presents several technical challenges: (1) Mesh sensitivity: ADM / ALM requires the application of distributed forces into the flow field, making it highly sensitive to mesh resolution, mesh quality, and force distribution methods. Mesh variations can lead to significant differences in results. (2) Numerical stability issues: Introducing source terms into the flow field can cause numerical instability, especially with large time steps or poor mesh quality, which can easily lead to divergence or oscillations. (3) Complex boundary condition handling: The coupling between the actuator disk / actuator line and the flow field boundary conditions is complex and may affect computational convergence. (4) Difficulty in force distribution parameterization: The force distribution function in ADM / ALM needs to be parameterized according to the specific fan model, lacking universality. In the integrated design process of floating wind turbines, there exists a critical technical window: the conceptual design phase and the preliminary design phase. During this phase, designers face the following typical challenges: Design iteration requirements: (1) It is necessary to quickly assess the impact of different impeller specifications (power level, impeller diameter, hub height) on the performance of the floating platform; (2) It is necessary to quickly match the compatibility of the floating platform type (semi-submersible, tension leg, barge type) with the impeller specifications; (3) It is necessary to quickly optimize the platform geometry parameters (buoy size, column spacing, draft) to adapt to different impellers.

[0005] Parameter uncertainties: (1) The detailed aerodynamic parameters of the impeller (blade airfoil, twist angle distribution, chord length distribution) have not been fully determined; (2) The impeller control strategy (pitch control, yaw control, power control) is still being optimized; (3) The impeller-platform coupling effect (aerodynamic-hydraulic-structural coupling) needs to be quickly evaluated.

[0006] Time pressure: (1) The project is in a tight timeframe in the early stages, requiring the comparison of multiple options within a limited time; (2) Design changes are frequent, requiring rapid response to parameter adjustments; (3) Cost control requirements necessitate a rapid assessment of the economics of different configurations.

[0007] The limitations of existing technologies in the early and mid-stage design phases are as follows: For full-scale CFD simulation, on the one hand, geometric modeling is complex, requiring remodeling of blade geometry for each impeller parameter adjustment. On the other hand, mesh generation is complex, with different impeller specifications requiring different meshing strategies, resulting in excessively high computational costs. Extensive comparison of various schemes is necessary in the early stages, making full-scale simulation prohibitively expensive. For actuation disk / actuator line models, the inherent force distribution parameterization requirements are complex, necessitating determination of the force distribution function based on the specific impeller model. With the impeller model undetermined in the early stages, detailed impeller aerodynamic models, including blade geometry and airfoil data, are difficult to establish. For potential flow-based coupled models (such as OpenFAST), while computationally efficient, significant limitations exist in the early design phase. These limitations include limited ability to handle hydrodynamic nonlinearities; potential flow theory struggles to accurately handle large-amplitude motions, wave breaking, vortex shedding, and other nonlinear hydrodynamic phenomena, which are crucial in floating wind turbine design. Furthermore, due to the neglect of viscous effects, accurate prediction of the vortex structure, boundary layer effects, and viscous drag around the platform is impossible, affecting the accuracy of platform motion response prediction. Additionally, these methods are highly parameter-dependent, requiring extensive empirical parameter calibration. Therefore, for the early stages of integrated design of floating wind turbines, the technical solution based on viscous flow solvers has technical advantages, including the ability to handle nonlinear hydrodynamic problems: it can accurately handle complex hydrodynamic phenomena such as large amplitude motion, wave nonlinearity, and eddy shedding, providing more reliable performance predictions for platform design, especially for verification under extreme conditions, thus improving the applicability and reliability in the early design stage.

[0008] Therefore, there is an urgent need for a technical solution that can achieve high-fidelity coupled simulation using only the most basic performance parameters (rated power, impeller diameter, cut-in and cut-out wind velocities, and thrust coefficient at key wind speed points) even when the detailed impeller parameters are not yet determined. This would support rapid scheme comparison and design iteration in the early stages of integrated design. The solution should be based on a viscous flow solver, capable of accurately handling nonlinear hydrodynamic problems while maintaining high computational efficiency. Summary of the Invention

[0009] This invention provides a high-efficiency, high-fidelity coupled simulation method and system for floating wind turbines. By simplifying the aerodynamic effects of the impeller through virtual forces, it focuses on the response prediction of the entire machine, significantly reducing computational costs and improving numerical stability while ensuring coupling fidelity.

[0010] The technical solution of the present invention: A high-efficiency, high-fidelity coupled simulation method for floating wind turbines includes the following steps: A1. Determine the current position of the preset wheel hub reference point in global coordinates based on the rigid body pose; A2. In the parallel partitioned grid, determine the grid cell where the current position is located and its owning process, forming a unique owner-cell correspondence; A3. The process reads the velocity vector within the unit, constructs the rigid body motion velocity at the hub, and calculates the relative wind speed. A4. Based on quaternion attitude, the reference axis is transformed into the current axis direction of the wind turbine, and the relative wind speed is projected as the axial component. A5. Based on the wind speed-thrust discrete curve, the axial wind speed is quickly interpolated to obtain the thrust scalar and generate the thrust vector along the current axis of the wind turbine. A6. The torque is obtained by cross product of the radius vector of the hub relative to the center of rotation and the thrust vector; A7. Combine the thrust, torque, fluid integral force, and gravity as the external force and torque in the six-degree-of-freedom rigid body dynamics equation to advance the rigid body's posture. A8. Perform global consistency synchronization of the thrust and the torque in a parallel environment.

[0011] The relative wind speed is calculated as follows: , in, Let v be the velocity vector within the unit, v be the translational velocity of the rigid body, ω be the angular velocity of the rigid body, and r be the radius vector of the hub position relative to the center of rotation. This represents the current position of the wheel hub reference point in global coordinates. Let the position of the rotation center be in global coordinates. This refers to relative wind speed.

[0012] The current axial direction of the wind turbine is obtained by performing a quaternion attitude transformation on the reference axis of the machine coordinate system, where the reference axis is the x-axis of the machine. The axial component is calculated using the following formula: , in, Let Q be the unit vector of the reference axis, and let Q be the quaternion of the rigid body's current attitude, used to transform the reference axis from the slave system to the global system. As the reference axis of the machine system, This is a scalar representation of the axial component of the relative wind speed along the current axis. The inflow velocity used for interpolation.

[0013] The fast interpolation in step A5 includes: A51. First, a threshold determination is made for the axial wind speed based on the cut-in wind speed and the cut-out wind speed. If the axial wind speed is lower than the cut-in wind speed or higher than the cut-out wind speed, the thrust scalar is 0. A52. If the axial wind speed is between the cut-in wind speed and the cut-out wind speed, then the adjacent discrete points on the wind speed-thrust discrete curve are located by binary search, and then the thrust scalar is generated by linear interpolation.

[0014] Before interpolation, a threshold determination is made for the axial wind speed based on the cut-in wind speed and the cut-out wind speed to suppress the thrust in the non-working range.

[0015] Determining the correspondence between the owner and the unit in step A2 includes: A21. Each parallel process performs a local cell lookup for the current position of the hub reference point; A22. Summarize the search results of each process and select the owner process through reduction; A23. Using the owner process as the root, broadcast the owner process information and the corresponding grid cell information to all parallel processes.

[0016] When the external force and external torque are combined in step A7, it includes: C1. Fluid integral force and integral torque; C2. The equivalent aerodynamic load consisting of the thrust vector and its moment; C3. Gravity and the torque caused by gravity.

[0017] The thrust vector is along the current axis of the wind turbine, and the torque is the cross product of the radius vector and the thrust vector.

[0018] A partially coupled simulation system for a floating wind turbine, characterized in that it includes: The hub position determination module is used to calculate the global position of the hub based on the rigid body pose. The parallel point positioning module is used to determine the owner process and the corresponding grid cell in the parallel partitioned grid, forming a unique owner-cell correspondence. The field value reading and relative wind speed construction module is used to read the flow velocity vector in the corresponding grid cell in the owner process, construct the rigid body motion velocity at the hub and calculate the relative wind speed. The axis determination and decomposition module is used to obtain the current axis direction of the wind turbine through quaternion attitude transformation and project the relative wind speed into an axial component. The thrust interpolation module is used to quickly interpolate the axial wind speed corresponding to the axial component based on the wind speed-thrust discrete curve, obtain the thrust scalar, and generate the thrust vector along the current axis direction of the wind turbine. The external load generation and synchronization module is used to obtain the torque based on the cross product of the radius vector of the hub relative to the rotation center and the thrust vector, and to perform global consistency synchronization of the thrust and the torque in a parallel environment; The six-degree-of-freedom dynamics module is used to combine the thrust, the torque, the fluid integral force, the fluid integral torque, and the gravity and the torque caused by gravity into the external forces and external torques of the six-degree-of-freedom rigid body dynamics equations to propel the rigid body's posture.

[0019] The parallel point localization module includes a local search unit, a global reduction unit, and a broadcast unit; The local lookup unit is used by each parallel process to perform a local unit lookup. The global reduction unit is used to summarize the search results of each process and select the owner process through a reduction algorithm; The broadcast unit is used to broadcast owner and unit information to all parallel processes with the owner process as the root.

[0020] The thrust interpolation module includes a threshold determination unit, a binary search unit, and a linear interpolation unit; The threshold determination unit is used to determine the threshold of axial wind speed based on the cut-in wind speed and the cut-out wind speed. The binary search unit is used to locate adjacent discrete points on the wind speed-thrust discrete curve for axial wind speeds between the cut-in wind speed and the cut-out wind speed. The linear interpolation unit is used to generate a thrust scalar based on the adjacent discrete points through linear interpolation, and to generate a thrust vector along the axial direction.

[0021] The beneficial effects of this invention are: (1) By using the link of “hub point wind intake + velocity correction + axial projection + curve interpolation + external load injection”, the wind turbine thrust channel can be reconstructed without applying distributed force in the flow field, which significantly improves coupling fidelity and numerical stability. (2) The owner-root broadcast parallel consistency mechanism is adopted to ensure that the single value source is injected in a consistent manner across the entire field, and to avoid multiple injections or missed injections; (3) Applying equivalent aerodynamic loads to the 6DoF equations at the point load / moment level has lower computational cost compared to explicit body mesh and actuation distribution model, is more suitable for integrated design in the early stage, and is easy to deploy and expand in engineering. (4) Minimize parameter requirements: Only basic specification parameters, operating parameters and key performance parameters are required. No detailed blade geometry information, control strategies or complex aerodynamic models are needed. It is particularly suitable for rapid scheme comparison in the early stage of integrated design. (5) Design iteration friendly: It supports rapid evaluation of the impact of different impeller specifications, platform types and geometric parameters. Parameter adjustment is simple, only the wind speed-thrust curve needs to be modified; it is suitable for comparison of a large number of schemes and design iteration. (6) Lowering the design threshold: High-fidelity coupled simulation can be performed even before the detailed parameters of the impeller are determined, providing a scientific basis for impeller selection, platform matching and concept verification, and significantly reducing the technical threshold and computational cost of the early design. (7) Advantages of viscous flow solution: Based on the solution of viscous flow equation, it can accurately handle nonlinear hydrodynamic problems, viscous effects and complex geometry. Compared with potential flow model, it has higher prediction accuracy and physical reality, and is particularly suitable for performance evaluation in the early design stage of floating wind turbine. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the external load generation process.

[0023] Figure 2 This is a schematic diagram of the virtual thrust generation process.

[0024] Figure 3 Set up logic code for real-time positioning and grid movement of the wind turbine's rotation center.

[0025] Figure 4 The logic code for determining the partition thread where the parallel point is located and the rotation center of the wind turbine is located is provided.

[0026] Figure 5 This is the logic code for 6DoF external load synthesis and pose propagation.

[0027] Figure 6 Update the logic code for mesh displacement.

[0028] Figure 7 Set up the logic code for the parallel point localization algorithm.

[0029] Figure 8 Set up the logic code for the thrust interpolation algorithm. Detailed Implementation

[0030] Example 1: Calculation Case 1.1 System Parameter Settings In one embodiment, taking a 20MW floating wind turbine as an example, the system parameters are set as follows: (1) Wind turbine parameters: rated power 20MW, hub height 155.468m, impeller diameter 260m, cut-in wind speed 3m / s, cut-out wind speed 25m / s; (2) Platform parameters: semi-submersible platform, total mass about 13,866,269kg, rotational inertia matrix is ​​a diagonal matrix, principal moments of inertia are Ixx=17633770730 kg·m², Iyy=17633771612 kg·m², Izz=34605628619 kg·m²; (3) Mesh parameters: computational domain size 1300m×780m×500m, total number of meshes about 3 million, number of parallel partitions 64; (4) Time parameters: total simulation duration 500s, time step 0.02s.

[0031] 1.2 Real-time positioning of the fan rotation center and grid motion setting Based on the construction process of Foam::sixDoFRigidBodyMotionSolver, the initial position parameters of the wheel hub and the relevant parameters of mesh motion are set. The scaling field of the mesh motion is implemented through OpenFOAM's native distance field calculation and cosine mapping. The specific code logic is as follows: Figure 3 As shown: Among them, the inner distance di=15m and the outer distance do=150m are used to control the smooth propagation range of the mesh motion.

[0032] 1.3 Parallel point location and determination of the partition thread where the wind turbine rotation center is located The function `findCellParallel(const point&)` is provided to locate a hub point in a partitioned mesh, determining the current cell and its owner process, and generating a unique "owner-cell" tuple. The specific code logic is as follows: Figure 4 As shown: This process uses global list reduction (such as maxEqOp) <label>The algorithm employs root broadcasting to ensure consistency and single-source injection in a parallel environment. This parallel point localization algorithm can effectively handle point localization problems in a distributed grid environment.

[0033] 1.4 Relative wind speed structure and axial projection Calculate the relative wind speed at the wheel hub: Subsequently, the reference axis is transformed to the current axis direction based on the quaternion orientation, and the axial component is calculated. .

[0034] 1.5 Wind speed-thrust curve interpolation and external load generation.

[0035] Based on the preset wind speed-thrust discrete table, The system employs binary search to locate adjacent points and performs linear interpolation to obtain the thrust scalar (T). The thrust vector and torque interpolation process are constructed on the owner and then synchronized to all processes via parallel broadcast for subsequent synthesis.

[0036] 1.6 6DoF Exogenous Load Synthesis and Pose Propulsion The fluid integral force and torque are obtained using the `forces` object, and synthesized with the equivalent thrust and torque and gravity term. This result is then input into the motion update routine to determine the propulsion pose. The specific code logic is as follows: Figure 5 As shown: 1.7 Mesh Displacement Update The mesh point displacement field is obtained through `motion_.transform(points0(), scale_)`, and point constraints are applied to ensure boundary and geometric consistency. The specific code logic is as follows: Figure 6 As shown: Example 2: Detailed Algorithm Flow: 2.1 Main Algorithm Flow The core algorithm flow of this patent is as follows: Step 1: Initialization Read the fan parameters (rated power, impeller diameter, cut-in / cut-out velocity, etc.); read the platform parameters (mass, moment of inertia, center of gravity position, etc.). Read environmental parameters (wind speed, wave height, period, etc.); Initialize the grid scaling field and parallel communication settings.

[0037] Step 2: Time Step Loop For each time step t, perform the following operations: Step 2.1: Pose Update: Calculate the current wheel hub position based on the pose of the previous moment.

[0038] Step 2.2: Parallel point localization: The function `findCellParallel(\(\mathbf{x}_{\text{hub}}(t)\))` is called; each process performs a local cell search: `localCell = mesh().findCell(\(\mathbf{x}_{\text{hub}}(t)\))`, and then the results from each process are summarized: `Pstream::listCombineGather(allFound, maxEqOp`. <label>Finally, the owner process is determined and the result is broadcast: Pstream::broadcast(result, ownerProc).

[0039] Step 2.3: Relative wind speed calculation: The owner process reads the flow velocity within the unit and obtains the floating body's movement speed, and then calculates the relative wind speed.

[0040] Step 2.4: Axial Projection: Obtain the current quaternion pose, then calculate the current axis direction, and finally calculate the axial component.

[0041] Step 2.5: Thrust interpolation: Perform binary search to locate neighboring points: binarySearch(thrustTable, U_parallel); calculate the thrust using linear interpolation and construct the thrust vector.

[0042] Step 2.6: Torque Calculation: Calculate the radius vector of the hub relative to the center of rotation, and then calculate the torque.

[0043] Step 2.7: Payload Synthesis and Broadcast: Synthesize the external payload in the owner process and then broadcast the payload to all processes: Pstream::broadcast(F_total, ownerProc).

[0044] Step 2.8: 6DoF Update: Call the motion update function: motion_.update(F_total, M_total,dt) to update the rigid body pose, velocity, and acceleration.

[0045] 2.2 Key Algorithm Implementation Details 2.2.1 Parallel point localization algorithm, the specific code logic is as follows: Figure 7 As shown.

[0046] 2.2.2 Thrust interpolation algorithm, the specific code logic is as follows: Figure 8 As shown.< / label> < / label>

Claims

1. A high-efficiency, high-fidelity coupled simulation method for floating wind turbines, comprising the following steps: A1. Determine the current position of the preset wheel hub reference point in global coordinates based on the rigid body pose; A2. In the parallel partitioned grid, determine the grid cell where the current position is located and its owning process, forming a unique owner-cell correspondence; A3. The process reads the velocity vector within the unit, constructs the rigid body motion velocity at the hub, and calculates the relative wind speed. A4. Based on quaternion attitude, the reference axis is transformed into the current axis direction of the wind turbine, and the relative wind speed is projected as the axial component. A5. Based on the wind speed-thrust discrete curve, the axial wind speed is quickly interpolated to obtain the thrust scalar and generate the thrust vector along the current axis of the wind turbine. A6. The torque is obtained by cross product of the radius vector of the hub relative to the center of rotation and the thrust vector; A7. Combine the thrust, torque, fluid integral force, and gravity as the external force and torque in the six-degree-of-freedom rigid body dynamics equation to advance the rigid body's posture. A8. Perform global consistency synchronization of the thrust and the torque in a parallel environment.

2. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, The relative wind speed is calculated as follows: , in, Let v be the velocity vector within the unit, v be the translational velocity of the rigid body, ω be the angular velocity of the rigid body, and r be the radius vector of the hub position relative to the center of rotation. This represents the current position of the wheel hub reference point in global coordinates. Let the position of the rotation center be in global coordinates. This refers to relative wind speed.

3. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, The current axial direction of the wind turbine is obtained by performing a quaternion attitude transformation on the reference axis of the machine coordinate system, where the reference axis is the x-axis of the machine. The axial component is calculated using the following formula: ， in, Let Q be the unit vector of the reference axis, and let Q be the quaternion of the rigid body's current attitude, used to transform the reference axis from the slave system to the global system. As the reference axis of the machine system, This is a scalar representation of the axial component of the relative wind speed along the current axis. The inflow velocity used for interpolation.

4. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, The fast interpolation in step A5 includes: A51. First, a threshold determination is made for the axial wind speed based on the cut-in wind speed and the cut-out wind speed. If the axial wind speed is lower than the cut-in wind speed or higher than the cut-out wind speed, the thrust scalar is 0. A52. If the axial wind speed is between the cut-in wind speed and the cut-out wind speed, then the adjacent discrete points on the wind speed-thrust discrete curve are located by binary search, and then the thrust scalar is generated by linear interpolation.

5. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, Before interpolation, a threshold determination is made for the axial wind speed based on the cut-in wind speed and the cut-out wind speed to suppress the thrust in the non-working range.

6. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, Determining the correspondence between the owner and the unit in step A2 includes: A21. Each parallel process performs a local cell lookup for the current position of the hub reference point; A22. Summarize the search results of each process and select the owner process through reduction; A23. Using the owner process as the root, broadcast the owner process information and the corresponding grid cell information to all parallel processes.

7. The high-efficiency, high-fidelity coupled simulation method for floating wind turbines according to claim 1, characterized in that, When the external force and external torque are combined in step A7, it includes: C1. Fluid integral force and integral torque; C2. The equivalent aerodynamic load consisting of the thrust vector and its moment; C3. Gravity and the torque caused by gravity. The thrust vector is along the current axis of the wind turbine, and the torque is the cross product of the radius vector and the thrust vector.

8. A partially coupled simulation system for a floating wind turbine used to implement the method of any one of claims 1 to 8, characterized in that, include: The hub position determination module is used to calculate the global position of the hub based on the rigid body pose. The parallel point positioning module is used to determine the owner process and the corresponding grid cell in the parallel partitioned grid, forming a unique owner-cell correspondence. The field value reading and relative wind speed construction module is used to read the flow velocity vector in the corresponding grid cell in the owner process, construct the rigid body motion velocity at the hub and calculate the relative wind speed. The axial determination and decomposition module is used to obtain the current axial direction of the wind turbine through quaternion attitude transformation and project the relative wind speed into an axial component. The thrust interpolation module is used to quickly interpolate the axial wind speed corresponding to the axial component based on the wind speed-thrust discrete curve, obtain the thrust scalar, and generate the thrust vector along the current axis direction of the wind turbine. The external load generation and synchronization module is used to obtain the torque based on the cross product of the radius vector of the hub relative to the rotation center and the thrust vector, and to perform global consistency synchronization of the thrust and the torque in a parallel environment; The six-degree-of-freedom dynamics module is used to combine the thrust, the torque, the fluid integral force, the fluid integral torque, and the gravity and the torque caused by gravity into the external forces and external torques of the six-degree-of-freedom rigid body dynamics equations to propel the rigid body's posture.

9. The partially coupled simulation system for a floating wind turbine according to claim 8, characterized in that, The parallel point localization module includes a local search unit, a global reduction unit, and a broadcast unit; The local lookup unit is used by each parallel process to perform a local unit lookup. The global reduction unit is used to summarize the search results of each process and select the owner process through a reduction algorithm; The broadcast unit is used to broadcast owner and unit information to all parallel processes with the owner process as the root.

10. The partially coupled simulation system for a floating wind turbine according to claim 8, characterized in that, The thrust interpolation module includes a threshold determination unit, a binary search unit, and a linear interpolation unit; The threshold determination unit is used to determine the threshold of axial wind speed based on the cut-in wind speed and the cut-out wind speed. The binary search unit is used to locate adjacent discrete points on the wind speed-thrust discrete curve for axial wind speeds between the cut-in wind speed and the cut-out wind speed. The linear interpolation unit is used to generate a thrust scalar based on the adjacent discrete points through linear interpolation, and to generate a thrust vector along the axial direction.

Images