A multi-grid fusion method for numerical simulation of wind turbine wake over complex terrain

Abstract

This invention provides a multi-grid fusion method for numerical simulation of wind turbine wakes in complex terrain. It fuses Cartesian grids of complex terrain, structured or unstructured hybrid grids of the wind turbine region, and anisotropic O-type structured grids of the wake region. By fusing multiple grid types with different topologies and using overlapping grid technology for assembly and interpolation calculations, this method significantly improves wake vortex analysis efficiency and controls computational costs while ensuring the accuracy of wind turbine aerodynamic loads and automated terrain geometry modeling. This enhances the accuracy and efficiency of numerical simulation of wind field wakes.

Description

Technical Field

[0001] This invention relates to the technical field of wind turbine engineering calculations, specifically a multi-grid fusion method for numerical simulation of wind turbine wakes in complex terrain. Background Technology

[0002] As a core component of the clean energy system, the key to the large-scale development of wind energy lies in the efficient operation of wind farms. Power generation, as a core indicator directly reflecting the operational efficiency and economic benefits of wind farms, requires accurate assessment and is a crucial prerequisite for improving overall power generation efficiency. In the assessment of wind farm power generation, the wake effect of wind turbines is a critical and unavoidable influencing factor.

[0003] The wake effect of wind turbines specifically refers to the disturbance of airflow by the rotating blades of upstream wind turbines during operation, which in turn alters the airflow state downstream, leading to a significant decrease in downstream wind speed and a substantial increase in turbulence intensity, ultimately resulting in a decline in the power generation efficiency of downstream wind turbines. Therefore, accurately assessing the wake effect is of great value. On the one hand, it can effectively improve the prediction accuracy of overall power generation; on the other hand, it can provide solid theoretical support for wind turbine micro-situation and cluster control strategies (such as yaw coordination), thus becoming one of the core technical challenges that need to be overcome in the wind farm design phase.

[0004] Currently, wind farms are rapidly developing towards larger scale, clustering, and greater complexity. As key equipment for wind energy conversion, the layout density and scale of wind turbines are constantly increasing, which further expands the scope and significantly enhances the impact of wake effects on power generation efficiency. This intensified wake effect will more profoundly alter the inflow conditions of downstream wind turbines, thus posing a greater challenge to the accurate assessment of wind farm power generation and the improvement of overall power generation efficiency.

[0005] Currently, research on wind turbine wake effects mainly relies on numerical calculation methods, including empirical models (Park models), actuation-based methods, and high-precision simulations using Computational Fluid Dynamics (CFD). CFD high-precision simulation methods, based on a grid discretization scheme for wind turbine geometry analysis, can completely capture the aerodynamic characteristics and wake vortex evolution process of the blade surface, making them the best choice for accurately simulating wind turbine wakes.

[0006] The foundation of high-precision CFD methods lies in grid discretization techniques. Currently, high-precision numerical simulations of complex wind field terrains typically rely on overlapping grid techniques to discretize different components. The specific approach is as follows:

[0007] (1) Mesh discretization of complex terrain components: Due to the complex terrain geometry, Cartesian meshes are automatically generated, and the mesh is densified in the near-ground region to approximate the real terrain contour; Based on the assumption of "hub height flow dominance" (>100m), atmospheric boundary layer inflow conditions are introduced, eliminating the need to generate body-fitted boundary layer meshes and simplifying the terrain discretization process;

[0008] (2) Mesh discretization of wind turbine components: In order to accurately capture the geometric details of the blades and the key boundary layer flow, wind turbines (especially the blade area) usually use structured meshes or hybrid unstructured meshes (such as hexahedral boundary layer meshes + tetrahedral outer meshes) to generate body-fitting meshes, ensuring the geometric fit accuracy of the blades and accurately simulating the influence of boundary layer flow on the aerodynamic loads on the blade surface.

[0009] (3) Wake region mesh discretization: In order to capture the influence of the wake generated by the wind turbine on the downstream flow field and the wind turbine, the Cartesian background mesh of the above-mentioned terrain is globally isotropically refined in the estimated wind turbine wake region in an attempt to improve the resolution of the wake vortex.

[0010] To capture the wake vortex structure of wind turbines, current technologies necessitate globally and isotropically refining the Cartesian mesh in the wake region. While this "simple and crude" refinement method can improve resolution to some extent, its inherent global and isotropic characteristics are severely mismatched with the geometric properties of the target physical phenomenon (axially stretched, helical, slender vortices). This leads to two core problems: an explosive increase in mesh size and low analytical efficiency. Consequently, the computational cost of high-precision, large-scale wind farm wake simulation becomes extremely high, even making it difficult to achieve. Summary of the Invention

[0011] The purpose of this invention is to address the problem that existing technologies using global, isotropic Cartesian mesh refinement methods to capture wind turbine wake structures are severely mismatched with the geometric characteristics of the target physical phenomena. This leads to an explosive increase in mesh size, low analytical efficiency, and ultimately, extremely high or even impossible computational costs for high-precision, large-scale wind farm wake simulation. This invention significantly improves wake analysis efficiency while controlling computational costs, thereby enhancing the accuracy and efficiency of wind farm wake numerical simulation. Specifically, this invention provides a multi-mesh fusion method for numerical simulation of wind turbine wakes in complex terrain.

[0012] This invention proposes a multi-grid fusion method for numerical simulation of wind turbine wakes in complex terrain, comprising the following steps:

[0013] Step 1: Perform surface geometry modeling on the complex terrain to generate a continuous terrain surface geometry model. Discretize the terrain surface geometry model into triangular patches to generate a terrain surface mesh. Generate an outer envelope cube of the terrain surface mesh. Expand the outer envelope cube to obtain an extended computational domain. Determine the initial mesh size and generate a uniform Cartesian mesh. Based on the positional relationship between the mesh cells of the uniform Cartesian mesh and the triangular patches, refine the uniform Cartesian mesh to generate a Cartesian mesh for the complex terrain region in the extended computational domain.

[0014] Step 2: Construct the geometric model of the wind turbine blade, discretize the geometric model of the wind turbine blade into a wind turbine blade mesh, generate the surface mesh of the wind turbine blade, fill the surface mesh to generate a volume mesh, and generate an anisotropic structured boundary layer mesh on the surface of the wind turbine blade. The surface mesh, volume mesh, and boundary layer mesh of the wind turbine blade constitute the structured or unstructured hybrid mesh of the wind turbine region.

[0015] Step 3: Generate anisotropic wake region O-type structure mesh. Construct a wake volume by geometrically abstracting and extracting feature lines from the wind turbine wake region. Transform the complex wind turbine wake flow field into a geometric model. Construct O-type topological blocks based on the wake volume. The topological structure of the O-type topological blocks is a closed O-type mesh in cross-section. Refine the O-type mesh to obtain the wake region O-type structure mesh.

[0016] Step 4: Overlay and assemble the Cartesian mesh of the complex terrain region, the structured or unstructured hybrid mesh of the wind turbine region, and the O-type structured mesh of the wake region to obtain a unified computational domain.

[0017] Preferably, step 1: performing surface geometry modeling on complex terrain to generate a continuous terrain surface geometry model, discretizing the terrain surface geometry model into triangular patches to generate a terrain surface mesh, generating an outer envelope cube of the terrain surface mesh, expanding the outer envelope cube to obtain an expanded computational domain, determining the initial mesh size, generating a uniform Cartesian mesh, and refining the uniform Cartesian mesh based on the positional relationship between the mesh cells of the uniform Cartesian mesh and the triangular patches to generate a Cartesian mesh for the complex terrain region of the expanded computational domain, includes the following steps:

[0018] Step 1.1: Perform geometric modeling on the complex terrain surface to generate a continuous terrain surface geometric model;

[0019] Step 1.2: Discretize the continuous terrain surface into triangular patches. Determine the patch position by recording the three-dimensional coordinates of the three vertices of each triangular patch, and record one normal vector to determine the patch orientation, thereby generating a terrain surface mesh.

[0020] Step 1.3: Construct an outer envelope cube based on the terrain surface mesh to generate a three-dimensional computational domain. Expand the three-dimensional computational domain based on simulation requirements. Extend the range upward to 10 to 100 times the top of the atmospheric boundary layer and extend the range horizontally to 10 to 20 times the wind turbine spacing or 2 to 3 times the wind field scale to obtain the expanded computational domain.

[0021] Step 1.4: Construct an initial mesh framework covering the extended computational domain, and generate a uniform Cartesian mesh based on an initial mesh size of 30~50m;

[0022] Step 1.5: Identify the positional relationship between the grid cells of the uniform Cartesian grid and the triangular facets in the terrain surface grid using ray casting or scanning methods, and refine the cells of the uniform Cartesian grid based on the positional relationship;

[0023] Step 1.6: If the grid cells of the uniform Cartesian grid intersect with the triangular facets in the terrain surface grid or the distance is less than the first threshold, then the grid cells of the uniform Cartesian grid are densified until the size of the grid cells of the uniform Cartesian grid is less than the second threshold, so as to approximate the terrain geometry.

[0024] Preferably, in step 1.6: if a grid cell of the uniform Cartesian mesh intersects with a triangular facet in the terrain surface mesh or the distance is less than a first threshold, then the grid cells of the uniform Cartesian mesh are densified until the size of the grid cells of the uniform Cartesian mesh is less than a second threshold, in order to approximate the terrain geometry, wherein:

[0025] The first threshold is 10m, and the second threshold is 10~20m;

[0026] The encryption rule for the uniform Cartesian grid is to divide the grid cell of the uniform Cartesian grid in half along the x, y, and z directions respectively. If the encrypted grid cell does not meet the condition that the grid size is less than the second threshold, the grid cell is further divided until the size of the grid cell is less than the second threshold.

[0027] Where x represents the horizontal east-west coordinate of the grid cell, y represents the horizontal north-south coordinate of the grid cell, and z represents the vertical height coordinate of the grid cell.

[0028] Preferably, step 2: constructing a geometric model of the wind turbine blade, discretizing the geometric model of the wind turbine blade into a wind turbine blade mesh, generating a surface mesh of the wind turbine blade, filling the surface mesh to generate a volume mesh, and generating an anisotropic structured boundary layer mesh on the surface of the wind turbine blade. The surface mesh, volume mesh, and boundary layer mesh of the wind turbine blade constitute a structured or unstructured hybrid mesh of the wind turbine region, including the following steps:

[0029] Step 2.1: Construct the initial geometric model of the wind turbine blade. Perform geometric cleaning and parameterization on the initial geometric model of the wind turbine blade. Use a non-uniform rational B-spline surface to describe the wind turbine blade with high precision to obtain the geometric model of the wind turbine blade.

[0030] Step 2.2: Based on the geometric model of the wind turbine blade, generate a two-dimensional wind turbine blade surface mesh. The mesh unit is a quadrilateral or triangular structure. Fill the wind turbine blade surface mesh to generate the wind turbine blade body mesh, and generate a boundary layer mesh on the wind turbine blade surface.

[0031] Preferably, step 2.2: Based on the geometric model of the wind turbine blade, a two-dimensional wind turbine blade surface mesh is generated, with the mesh cells being quadrilateral or triangular structures. The wind turbine blade surface mesh is filled to generate a wind turbine blade body mesh, and a boundary layer mesh is generated on the surface of the wind turbine blade, including the following steps:

[0032] Step 2.2.1: Generate an anisotropic structured blade surface boundary layer mesh on the surface of the wind turbine blade, with the mesh lines forming an angle ≥85° with the normal direction of the wind turbine blade surface;

[0033] Step 2.2.2: Generate a boundary layer mesh of triangular prism shape at the connection points of each component of the wind turbine blade;

[0034] Step 2.2.3: Adjust the fineness of the blade surface boundary layer mesh and the connection position boundary layer mesh in steps 2.2.1 and 2.2.2. Adjust the thickness of the first boundary layer mesh according to the preset y⁺ value. Set the number of layers of the blade surface boundary layer mesh and the connection position boundary layer mesh to ≥30 layers and the thickness gradient ratio to ≤1.2 to form the boundary layer mesh of the wind turbine blade surface. Here, y⁺ is the ratio of the distance from the center of the first mesh to the wind turbine surface to the viscous length of the airflow.

[0035] Step 2.2.4: Generate a tetrahedral mesh, which smoothly transitions from the outer edge of the boundary layer mesh on the surface of the wind turbine blade to the far field, with a transition range of approximately 10 times the blade chord length.

[0036] Preferably, step 3: generating an anisotropic wake region O-type structure mesh, constructing a wake volume by geometrically abstracting and extracting feature lines from the wind turbine wake region, transforming the complex wind turbine wake flow field into a geometric model, constructing O-type topological blocks based on the wake volume, wherein the topological structure of the O-type topological blocks is a closed O-type mesh in cross-section, and refining the O-type mesh to obtain the wake region O-type structure mesh, includes the following steps:

[0037] Step 3.1: Geometrically abstract the wind turbine wake region, representing it as a cylinder starting from a certain distance upstream of the impeller and extending downstream. Extract feature lines from the wind turbine wake region, extracting the wake centerline and boundary shear layer to construct the wake volume.

[0038] Step 3.2: Based on the wake, construct an O-type topology block with the wind turbine's rotation axis as the central axis. The topology of the O-type topology block is a closed O-type mesh in cross-section. The diameter of the cylindrical computational domain of the O-type mesh is 10 times the diameter of the wind turbine rotor, and the length is 150 times the diameter of the wind turbine rotor. The distance from the rotational plane where the wind turbine is located to the inlet of the cylindrical computational domain is 60 times the diameter of the wind turbine rotor, and the distance from the outlet of the cylindrical computational domain is 90 times the diameter of the wind turbine rotor.

[0039] Step 3.3: Dense the O-type mesh radially within a range of 1.2 times the hub radius centered on the hub and within a range of 0.8 to 1.2 times the impeller radius, and axially denser the O-type mesh within a range of 5 times the impeller diameter upstream to 90 times the impeller diameter downstream of the rotation plane where the wind turbine is located.

[0040] Preferably, step 3.3: radially densifying the O-type mesh within a range of 1.2 times the hub radius centered on the hub and within a range of 0.8 to 1.2 times the impeller radius, and axially densifying the O-type mesh within a range of 5 times the impeller diameter upstream to 90 times the impeller diameter downstream of the rotation plane where the wind turbine is located, includes the following steps:

[0041] Step 3.3.1: Within a range of 1.2 times the hub radius centered on the hub, the O-type mesh is densified using a finer mesh of 0.2 times the hub diameter; within a range of 0.8 times the impeller radius to 1.2 times the impeller radius, the O-type mesh is densified using an ultrafine mesh of 0.1 times the blade tip chord length.

[0042] Step 3.3.2: Based on the flow velocity attenuation rate and vortex structure dissipation scale, the O-type grid is non-uniformly segmented and refined with the rotation plane where the wind turbine is located as the 0 point. The area from 5D upstream to 5D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.01D. The area from 5D to 20D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.02D to 0.05D. The area from 20D to 90D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.1D. Where D is the diameter of the wind turbine rotor.

[0043] Preferably, the O-type topology block can also be a C-type or HO hybrid topology block.

[0044] Preferably, step 4: Overlapping and assembling the Cartesian mesh of the complex terrain region, the structured or unstructured hybrid mesh of the wind turbine region, and the O-type structured mesh of the wake region to obtain a unified computational domain, including the following steps:

[0045] Step 4.1: Analyze the Cartesian grid of the complex terrain region, the structured or unstructured hybrid grid of the wind turbine region, and the O-type structured grid of the wake region, including the grid cell node coordinates, grid cell connectivity, and boundary condition markings, and evaluate the grid cell skewness, aspect ratio, and smoothness index.

[0046] Step 4.2: Determine the attributes of the mesh cells. For the Cartesian mesh of the complex terrain region and the structured or unstructured hybrid mesh of the wind turbine region, calculate the minimum distance d1 from the mesh cell in the overlapping area of ​​the Cartesian mesh of the complex terrain region and the structured or unstructured hybrid mesh of the wind turbine region to the wall of this component, and the minimum distance d2 from the mesh cell in the overlapping area to the wall of other components. If d1 is less than or equal to d2, the mesh cell in the overlapping area is determined to be a valid cell. If d1 is greater than d2, the mesh cell in the overlapping area is determined to be an invalid cell. The valid cells close to the invalid cells are determined to be interpolation cells. For the O-type structure mesh of the wake region, set the wall distance of the mesh cell of the O-type structure mesh of the wake region to 10 times the maximum chord length of the wind turbine blade.

[0047] This invention proposes a multi-grid fusion method for numerical simulation of wind turbine wakes in complex terrain. By fusing various grid types with different topologies, such as unstructured hybrid grids of wind turbines, Cartesian grids of terrain, and anisotropic O-type structured grids of the wake region, and by using overlapping grid technology to achieve assembly and interpolation calculations, this method aims to significantly improve the efficiency of wake vortex analysis and control computational costs while ensuring the accuracy of wind turbine aerodynamic loads and automated modeling of terrain geometry, thereby enhancing the accuracy and efficiency of numerical simulation of wind field wakes. Attached Figure Description

[0048] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description of the embodiments or prior art will be briefly introduced below.

[0049] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0050] Figure 2 This is a schematic diagram of the continuous curved surface geometric model of complex terrain according to the present invention.

[0051] Figure 3 This is a schematic diagram of the local encryption of the Cartesian grid in this invention.

[0052] Figure 4This is a schematic diagram of the unstructured hybrid grid in the wind turbine area of ​​the present invention.

[0053] Figure 5 This is a schematic diagram of the boundary layer mesh on the blade surface of the present invention.

[0054] Figure 6 This is a schematic diagram of the O-shaped structure mesh in the wake region of the present invention being axially densified near the impeller plane of the wind turbine.

[0055] Figure 7 This is a schematic diagram of the radial densification of the O-shaped structure mesh in the wake region of the present invention near the blade root and tip of the wind turbine blade. Detailed Implementation

[0056] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0057] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0058] Figure 1 This is a schematic diagram of a multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain, according to the present invention. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain proposed in this invention includes the following steps:

[0059] Step 1: Generate a Cartesian grid for complex terrain regions.

[0060] Includes the following steps:

[0061] Step 1.1: As Figure 2As shown, terrain surface geometry modeling is performed. Complex terrain includes mountains, hills, canyons, etc. The construction method of three-dimensional surface geometry is to use common CAD software and on-site survey data (such as UAV aerial survey, LiDAR laser scanning data) to restore the details of terrain undulation, slope, gullies, etc., and generate a continuous surface geometry model.

[0062] Step 1.2: Terrain Surface Discretization. The continuous surface is discretized into a finite number of triangular patches, generating a terrain surface mesh. The triangular patch is the smallest constituent unit of the terrain surface, representing a high-fidelity polygonal approximation of the original continuous surface. Each triangular patch's position is determined by recording the 3D coordinates of its three vertices, and its orientation is determined by recording one normal vector, facilitating subsequent Cartesian mesh identification of terrain interiors and exteriors. This step can be implemented by exporting an STL format file using CAD software.

[0063] Step 1.3: Construct the 3D computational domain and expand it according to simulation requirements. Based on the terrain surface mesh from Step 1.2, establish an outer bounding box, which completely encloses the entire terrain surface mesh using the smallest possible cube. The boundary coordinates of the cube are the maximum or minimum x, y, and z coordinates of the 3D model, where x represents the east-west horizontal coordinate, y represents the north-south horizontal coordinate, and z represents the vertical height coordinate. The boundary of the outer bounding cube forms the initial framework of the 3D computational domain.

[0064] Subsequently, the computational domain was expanded according to the simulation requirements. Since the wind turbine wake diffuses throughout the atmosphere, not just confined to the terrain surface, the range needed to be expanded outwards from the 3D model of the complex terrain. The atmospheric boundary layer is the core area of ​​wind turbine operation; it refers to the atmospheric region closest to the Earth's surface, most significantly affected by terrain and surface friction. The atmospheric boundary layer extends upwards from the ground, with a height of approximately 100-1000m, depending on the climate, weather, and surface type of the simulated area; its height is not a fixed value. For example, in plains areas on sunny days, the atmospheric boundary layer is about 200-300m, and may reach 500m on cloudy or windy days; in mountainous or complex terrain, the atmospheric boundary layer is thicker due to the disrupted airflow caused by topographical undulations, potentially reaching 500-1000m; in tropical regions, where surface heating is strong, the boundary layer can reach over 1000m; and in polar regions, the atmospheric boundary layer is thinner, about 100-200m.

[0065] The key to extending the atmospheric boundary layer is upward extension, because the wake of a wind turbine spreads upward, typically extending to 10 to 100 times the top of the atmospheric boundary layer. The top of the atmospheric boundary layer refers to the upper boundary of the atmospheric boundary layer. In the atmospheric region above the upper boundary, the influence of surface friction is basically eliminated, and airflow is mainly controlled by atmospheric circulation (such as the upper-level jet stream). Wind speed and direction are more stable and almost unaffected by terrain. If simulating only the short-term wake of a single wind turbine, such as the diffusion within 10 minutes, an extension range of 10 to 20 times the atmospheric boundary layer top is sufficient. For example, if the atmospheric boundary layer top is 300m, the computational domain should be extended upwards to 3000 to 6000m, which is enough for the wake to diffuse completely within the domain. If simulating the long-term wake of a large wind farm, such as the wake of dozens of wind turbines for more than 1 hour, an extension range of 50 to 100 times is required. For example, if the atmospheric boundary layer top is 500m, the computational domain needs to be extended upwards to 25000 to 50000m to ensure that the wakes of multiple wind turbines can still move completely within the computational domain without being truncated after being superimposed.

[0066] Meanwhile, the horizontal direction of the atmospheric boundary layer is appropriately extended according to the wind turbine spacing and wind field scale, usually by 10 to 20 times the wind turbine spacing or 2 to 3 times the wind field scale, to avoid the wake hitting the boundary of the computational domain and being truncated.

[0067] Step 1.4: Determine the initial mesh sizes Δx, Δy, and Δz, and generate a uniform Cartesian mesh.

[0068] Establish an initial mesh framework covering the entire computational domain. The initial mesh sizes Δx, Δy, and Δz represent the side lengths of the mesh in the x, y, and z directions, respectively, with commonly used engineering values ​​of 30-50m. If the initial mesh size is too large, such as 100m, small terrain undulations will be ignored; if it is too small, such as 10m, the total number of meshes will increase dramatically, causing computational lag.

[0069] Step 1.5: Identify the positional relationship between the Cartesian grid cells and the triangular facets in the terrain surface grid using ray casting or scanning methods.

[0070] This step is used to determine whether each Cartesian mesh touches the triangular mesh on the terrain surface, thus identifying the area requiring refinement. The ray casting method works by emitting a virtual ray from the center point of a mesh cell in any direction and counting the number of intersections between the ray and the triangular facets. An odd number of intersections indicates the mesh is inside the terrain (e.g., underground), an even number indicates it's outside (e.g., in the atmosphere), and a perfect touch indicates an intersection (half of the mesh is inside the terrain, half outside). The scanning method works by scanning the computational domain row by row in the x or y direction, recording the overlap between each mesh and the triangular facet, and distinguishing the positional relationship between the Cartesian mesh and the triangular facet, including interior, exterior, and intersection relationships.

[0071] Step 1.6: As Figure 3As shown, if a Cartesian grid cell intersects with a triangular facet of the terrain surface grid or the distance is less than a first threshold, the grid cell is densified until the grid size is less than a second threshold in order to approximate the terrain geometry.

[0072] This step refines the key mesh near the terrain, making the mesh fit the terrain undulations and improving simulation accuracy.

[0073] There are two triggering conditions for refining the mesh cells. The first triggering condition is that the Cartesian mesh cell intersects with the triangular facet of the terrain surface mesh (Cut-cell), meaning the mesh is cut by the terrain facet and its shape is no longer a complete cube. If this type of mesh is not refined, it will cause the terrain surface to appear jagged, such as a steep slope being cut into steps by a coarse mesh. The second triggering condition is that the distance between the Cartesian mesh cell and the triangular facet of the terrain surface mesh is less than the first threshold, usually 10m. At this time, even if the Cartesian mesh cell and the triangular facet of the terrain surface mesh do not directly intersect, the terrain still has a significant impact on the airflow within a 10m radius. For example, the airflow near a hillside will flow around it, and a fine mesh is needed to capture this change.

[0074] The encryption rule is as follows: A grid that meets the conditions is divided in half along the x, y, and z directions, resulting in eight sub-grids. The side length of each sub-grid is half the side length of the original grid, i.e., 25m. If the encrypted grid still does not meet the condition that the grid size is less than the second threshold, the grid is further divided until the grid size is less than the second threshold. The second threshold is set according to the terrain complexity; the more complex the terrain, the smaller the threshold, typically 10-20m.

[0075] Step 2: Generate a structured or unstructured hybrid mesh for the wind turbine area, such as... Figure 4 As shown.

[0076] This step is used to generate the mesh for the core region of the wind turbine, and specifically includes the following steps:

[0077] Step 2.1: Geometric Modeling of Wind Turbine Blades. This involves transforming the complex structure of the physical wind turbine into a high-precision, computer-recognizable geometric model.

[0078] Since the original CAD model of a wind turbine may contain redundant information (such as non-critical details like bolts and holes), this invention uses CAD software to perform geometric cleanup and parametric characterization on the wind turbine blades, a key component of the wind turbine system, generating a high-quality STL model based on the wind turbine blades. Geometric cleanup refers to deleting features that do not affect airflow simulation, avoiding unnecessary increases in the number of meshes. Parametric characterization refers to defining key dimensions using mathematical formulas, such as the variation of the blade's chord length and twist angle with the span, facilitating subsequent modifications. For example, when adjusting the blade shape, only the parameters need to be changed to update the model.

[0079] This invention employs non-uniform rational B-spline surfaces to accurately describe the geometric contours of complex aerodynamic shapes such as blades, providing accurate geometric definitions for subsequent high-quality mesh generation. Non-uniform rational B-spline surfaces can precisely describe complex curves and surfaces using only a few control points, offering smoother surfaces and lower errors compared to triangular patches.

[0080] Step 2.2: Discretization of wind turbine blade mesh. The geometric model is decomposed into a finite number of meshes, allowing the computer to calculate airflow motion through the meshes. This step is divided into three stages: surface mesh generation, volume mesh generation, and boundary layer mesh generation.

[0081] First, a surface mesh is generated on the wind turbine blades. A two-dimensional mesh is generated on the surface of the blades, and the mesh cells are usually quadrilateral or triangular structures.

[0082] Subsequently, a volume mesh is generated based on the surface mesh. The interior of the surface mesh is filled with a three-dimensional mesh. This step is the core of the airflow calculation. By solving the momentum equation and energy equation within each volume mesh cell, the computer can obtain parameters such as flow velocity and pressure.

[0083] Finally, as Figure 5 As shown, an anisotropic structured boundary layer mesh is generated on the surface of the wind turbine blade, ensuring that the mesh orthogonality is >85°, i.e., the angle between the mesh lines and the normal direction of the blade surface is ≥85° (close to perpendicular). If the orthogonality is poor (e.g., 60°), the mesh will be oblique, which will amplify the error during calculation and lead to deviations in the initial intensity of the wake. Simultaneously, triangular prism meshes are used locally in geometrically abrupt regions at component connection points, such as the connection between the blade and the hub, and the transition zone between different airfoil sections of the blade, to avoid distortion. Furthermore, the fineness of the boundary layer mesh is controlled by dynamically adjusting the first layer thickness according to a preset y⁺ value (approximately 1), setting the number of boundary layer mesh layers to ≥30 (thickness gradient ratio ≤1.2) to generate the boundary layer mesh. From the outer edge of the boundary layer to the far field, the transition range is approximately 10 times the blade chord length, and a tetrahedral mesh is generated for a smooth transition.

[0084] The physical meaning of y⁺ is: it measures the ratio of the distance from the center of the first layer of the grid on the wind turbine blade surface to the viscous length of the airflow, and is a core indicator for judging whether the grid on the wind turbine surface is sufficiently fine. For wind turbine simulation, y⁺≈1 means that the first layer of the grid can capture the region near the wind turbine surface where the flow velocity increases sharply from 0. The airflow frictional resistance in this region is the key source of blade stress and initial turbulence in the wake.

[0085] A boundary layer mesh with ≥30 layers means that the boundary layer is the transition region from the fan surface (where the velocity is 0) to the external free flow (where the velocity is close to the incoming velocity). A sufficient number of layers is needed to capture continuous changes in velocity. If the number of layers is too small, such as 10 layers, the velocity will jump, and the velocity difference between adjacent mesh layers will be too large, resulting in an underestimation of the turbulence intensity in the wake.

[0086] A thickness gradient ratio ≤ 1.2 means that the thickness ratio of two adjacent mesh layers is ≤ 1.2. For example, if the thickness of the first layer is 0.1 mm, the second layer is ≤ 0.12 mm, the third layer is ≤ 0.144 mm, and so on. If the gradient ratio is too large, such as 1.5, the mesh will suddenly become thicker, making it impossible to distinguish the fine velocity changes within the boundary layer.

[0087] The blade chord length is the distance from the leading edge to the trailing edge of the blade's largest cross-section, typically 2-5m, and 10 times the chord length is 20-50m. This is the interpolation boundary between the flow field near the wind turbine blade and the background grid. Beyond 10 times the chord length, the interpolation boundary is moved away from the blade surface, and the intensity of the flow field disturbance from the blade surface begins to decrease, which can improve the accuracy of flow field interpolation.

[0088] Step 3: Generate anisotropic O-type structure mesh for the wake region.

[0089] Because the wind turbine wake is a low-speed disturbance region formed after the airflow passes over the blades, it contains complex vortex structures (such as tip vortices and root vortices) and velocity gradients. These characteristics gradually evolve downstream from the wind turbine outlet, resulting in a clear near-field vortex structure and gradual dissipation in the far field. Ordinary meshes struggle to balance the fine structure of the wake core with the computational efficiency required for large-scale diffusion. In contrast, the O-shaped mesh can form concentric circular meshes centered on the wind turbine, tightly enclosing the wake core while smoothly transitioning to the far field, perfectly matching the characteristics of strong disturbance at the wake center and gradual decay outwards.

[0090] Step 3.1: Geometric abstraction and feature line extraction of the wake region to construct the wake volume. The complex wake flow field is transformed into a quantifiable geometric framework.

[0091] Based on flow field simulation results or empirical models, the wake region behind the blade is geometrically abstracted and represented as a cylinder extending downstream from a certain distance upstream of the impeller, preferably one impeller diameter. Core feature lines of this region, including the wake centerline and boundary shear layer, are extracted to define its principal axis direction and lateral expansion pattern, providing a geometric basis for constructing the topological framework of the O-type mesh.

[0092] Among them, the wake centerline refers to the axis of the cylinder, which determines the orientation of the grid and ensures that the grid extends along the wake development direction; the boundary shear layer refers to the boundary between the blade wake and the surrounding airflow, where the velocity gradient is the largest, and the radial boundary of the grid can be defined to ensure that the shear layer region has a sufficiently dense grid.

[0093] Step 3.2: Construct O-type topology blocks. O-type topology blocks are the basic mesh structure of the wake region. The specific construction method is as follows:

[0094] Around the abstract wake, an O-type topology block is constructed with the turbine shaft as the central axis. Alternatively, a C-type or HO hybrid topology block can be used. The topology of this block is a closed O-type mesh in cross-section, capable of high-quality encirclement of the wake core. The cylindrical computational domain of the O-type mesh has a diameter 10 times the turbine diameter (D) and a length of 150D. Using the turbine diameter (D) as a reference, the turbine's rotational plane is positioned at a specific location within the cylindrical computational domain of the O-type mesh. The turbine's location is 60D upstream; the distance from the turbine's rotational plane to the computational domain inlet (incoming flow direction) is 60 times the turbine diameter; the turbine's location is 90D downstream; and the distance from the turbine's rotational plane to the computational domain outlet (wake dissipation direction) is 90 times the turbine diameter.

[0095] Step 3.3: Anisotropic mesh refinement. To accurately capture high shear and strong vorticity regions, radial mesh refinement is required near the hub and blade tip, and axial mesh refinement is required before and after the rotation plane. Radial mesh refinement is perpendicular to the turbine axis (from center to outside), and axial mesh refinement is along the turbine axis (from upstream to downstream).

[0096] like Figure 7 When refining the mesh radially, the mesh is refined in the hub center and blade tip region: a fine mesh of 0.2 times the hub diameter is used in the range of 1.2 times the hub radius centered on the hub, and an ultrafine mesh of 0.1 times the blade tip chord length is used in the range of 0.8 times the impeller radius to 1.2 times the impeller radius.

[0097] like Figure 6 During axial mesh refinement, the axial direction is non-uniformly segmented based on the physical field gradient changes such as the flow velocity attenuation rate and vortex structure dissipation scale. Refinement is applied in the near-field region with a large gradient, while gradually thinning occurs in the far-field region with a small gradient, achieving an anisotropic mesh distribution. This is crucial for ensuring vortex core resolution. In the near-field region, with the rotation plane as the zero point, a high-resolution mesh is used from 5D upstream to 5D downstream (-5D to +5D) to capture vortex generation, such as an axial step size of 0.01D. In the transition region, the mesh is gradually relaxed from 5D downstream to 20D (+5D to +20D), such as increasing the axial step size from 0.02D to 0.05D. In the far-field region, a large mesh is used from 20D downstream to 90D (+20D to +90D) to handle turbulent diffusion, such as an axial step size of 0.1D.

[0098] Step 4: Overlap and assemble the topography, wind turbine, and wake O-shaped cylindrical grids to obtain a unified grid.

[0099] Step 4.1: Read in the mesh topology information and perform mesh quality inspection.

[0100] The mesh files of each component (such as CGNS, FLUENT MSH) are read in, including the Cartesian mesh of the terrain region in step 1, the wind turbine blade mesh in step 2, and the O-type structure mesh of the wake region in step 3. Their node coordinates, element connectivity, and boundary condition markers are analyzed. The boundary condition markers indicate which elements belong to the wall (blade surface, ground, etc.) or inlet / outlet, etc. Simultaneously, mesh assembly and quality checks are performed, evaluating indicators such as mesh skewness, aspect ratio, and smoothness to ensure they meet the requirements of the CFD solver.

[0101] Step 4.2: Determine the properties of the mesh elements. The properties of the mesh elements are determined based on the wall distance. Valid elements participate in the flow field calculation, while invalid elements do not. Interpolation elements are used for interpolating flow field variables between meshes of different components.

[0102] First, the properties of the mesh elements are determined based on the wall distance. The minimum distance d1 from the mesh elements of the Cartesian mesh in the complex terrain region (step 1) and the wind turbine blade mesh (step 2) to the wall of this component is calculated. Then, the minimum distance d2 from the mesh elements in the overlapping area of ​​the Cartesian mesh in the complex terrain region (step 1) and the wind turbine blade mesh (step 2) to the walls of other components is calculated. The sizes of d1 and d2 are compared. If d1 is less than or equal to d2, it is a valid element; if d1 is greater than d2, it is an invalid element. Valid elements closer to invalid elements are interpolated elements. Since the O-type structure mesh in the wake region itself has no walls, the wall distance of the O-type structure mesh elements in the wake region can be directly given as 10 times the maximum chord length of the wind turbine blade.

[0103] Step 4 ensures the reliability of the mesh itself through mesh quality inspection, resolves overlapping conflicts by judging wall distance, and finally integrates the three independent meshes from steps 1 to 3 into a logically unified computational domain. This process leverages the high precision advantages of each mesh in specific regions while avoiding the contradictions of overlapping calculations, providing a high-quality, conflict-free computational platform for subsequent wake numerical simulations.

[0104] In summary, the implementation results of the multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain proposed in this invention demonstrate that the design intent has been achieved.

[0105] The technical effects achieved by the multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain provided by this invention are as follows:

[0106] (1) Significantly reduce computational costs and improve simulation efficiency

[0107] By employing a region-specific customized meshing strategy, redundant computations caused by traditional isotropic meshing across the entire wake region are avoided. Instead, an anisotropic mesh structure with axial stretching and radial refinement is used in the wake region to precisely match the geometric characteristics of the tip vortex. This significantly reduces unnecessary mesh cells while maintaining resolution, allowing computational resources to be concentrated in key physical regions, thereby significantly reducing computational load and shortening simulation time.

[0108] (2) Accurately analyze the wake flow characteristics and improve the accuracy of wake simulation.

[0109] In the wind turbine region, a hybrid mesh (boundary layer hexahedrons / triangular prisms + outer tetrahedrons) is used to accurately capture the aerodynamic loads on the blades; in the wake region, an anisotropic mesh explicitly matches the axial tensile characteristics of the vortex filaments, effectively maintaining the coherence and strength of the vortex core; complex terrain is rapidly generated and locally refined using Cartesian meshes. This multi-mesh fusion strategy achieves higher-precision numerical analysis on key issues such as wind turbine performance, wake vortex evolution, and terrain wind conditions, significantly improving the accuracy of wake simulation.

[0110] (3) Improve the automation level and engineering applicability of complex wind field modeling.

[0111] This method integrates meshes of different topological types using overlapping mesh technology. It retains the rapid generation advantage of Cartesian meshes in terrain modeling while enhancing adaptability through local optimization of structured / unstructured meshes. This approach supports modular mesh generation and allows for dynamic adjustment of mesh configurations in different regions according to simulation requirements, significantly improving the automation and engineering applicability of complex wind farm modeling.

[0112] It should be noted that, for those skilled in the art, the technical features in the above embodiments can be freely combined, and the resulting technical solutions also belong to the embodiments disclosed in this invention.

[0113] Furthermore, without departing from the principles of this invention, several improvements and modifications can be made to this invention, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain, characterized in that, Includes the following steps: Step 1: Perform surface geometry modeling on the complex terrain to generate a continuous terrain surface geometry model. Discretize the terrain surface geometry model into triangular patches to generate a terrain surface mesh. Generate an outer envelope cube of the terrain surface mesh. Expand the outer envelope cube to obtain an extended computational domain. Determine the initial mesh size and generate a uniform Cartesian mesh. Based on the positional relationship between the mesh cells of the uniform Cartesian mesh and the triangular patches, refine the uniform Cartesian mesh to generate a Cartesian mesh for the complex terrain region in the extended computational domain. Step 2: Construct the geometric model of the wind turbine blade, discretize the geometric model of the wind turbine blade into a wind turbine blade mesh, generate the surface mesh of the wind turbine blade, fill the surface mesh to generate a volume mesh, and generate an anisotropic structured boundary layer mesh on the surface of the wind turbine blade. The surface mesh, volume mesh, and boundary layer mesh of the wind turbine blade constitute the structured or unstructured hybrid mesh of the wind turbine region. Step 3: Generate anisotropic wake region O-type structure mesh. Construct a wake volume by geometrically abstracting and extracting feature lines from the wind turbine wake region. Transform the complex wind turbine wake flow field into a geometric model. Construct O-type topological blocks based on the wake volume. The topological structure of the O-type topological blocks is a closed O-type mesh in cross-section. Refine the O-type mesh to obtain the wake region O-type structure mesh. Step 4: Overlay and assemble the Cartesian mesh of the complex terrain region, the structured or unstructured hybrid mesh of the wind turbine region, and the O-type structured mesh of the wake region to obtain a unified computational domain, including the following steps: Step 4.1: Analyze the Cartesian grid of the complex terrain region, the structured or unstructured hybrid grid of the wind turbine region, and the O-type structured grid of the wake region, including the grid cell node coordinates, grid cell connectivity, and boundary condition markings, and evaluate the grid cell skewness, aspect ratio, and smoothness index. Step 4.2: Determine the attributes of the mesh cells. For the Cartesian mesh of the complex terrain region and the structured or unstructured hybrid mesh of the wind turbine region, calculate the minimum distance d1 from the mesh cell in the overlapping area of ​​the Cartesian mesh of the complex terrain region and the structured or unstructured hybrid mesh of the wind turbine region to the wall of this component, and the minimum distance d2 from the mesh cell in the overlapping area to the wall of other components. If d1 is less than or equal to d2, the mesh cell in the overlapping area is determined to be a valid cell. If d1 is greater than d2, the mesh cell in the overlapping area is determined to be an invalid cell. The valid cells close to the invalid cells are determined to be interpolation cells. For the O-type structure mesh of the wake region, set the wall distance of the mesh cell of the O-type structure mesh of the wake region to 10 times the maximum chord length of the wind turbine blade.

2. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 1, characterized in that, Step 1: Perform surface geometry modeling on complex terrain to generate a continuous terrain surface geometry model. Discretize the terrain surface geometry model into triangular facets to generate a terrain surface mesh. Generate an outer envelope cube of the terrain surface mesh. Expand the outer envelope cube to obtain an extended computational domain. Determine the initial mesh size and generate a uniform Cartesian mesh. Based on the positional relationship between the mesh cells of the uniform Cartesian mesh and the triangular facets, refine the uniform Cartesian mesh to generate a Cartesian mesh for the complex terrain region within the extended computational domain. This includes the following steps: Step 1.1: Perform geometric modeling on the complex terrain surface to generate a continuous terrain surface geometric model; Step 1.2: Discretize the continuous terrain surface into triangular patches. Determine the patch position by recording the three-dimensional coordinates of the three vertices of each triangular patch, and record one normal vector to determine the patch orientation, thereby generating a terrain surface mesh. Step 1.3: Construct an outer envelope cube based on the terrain surface mesh to generate a three-dimensional computational domain. Expand the three-dimensional computational domain based on simulation requirements. Extend the range upward to 10 to 100 times the top of the atmospheric boundary layer and extend the range horizontally to 10 to 20 times the wind turbine spacing or 2 to 3 times the wind field scale to obtain the expanded computational domain. Step 1.4: Construct an initial mesh framework covering the extended computational domain, and generate a uniform Cartesian mesh based on an initial mesh size of 30~50m; Step 1.5: Identify the positional relationship between the grid cells of the uniform Cartesian grid and the triangular facets in the terrain surface grid using ray casting or scanning methods, and refine the cells of the uniform Cartesian grid based on the positional relationship; Step 1.6: If the grid cells of the uniform Cartesian grid intersect with the triangular facets in the terrain surface grid or the distance is less than the first threshold, then the grid cells of the uniform Cartesian grid are densified until the size of the grid cells of the uniform Cartesian grid is less than the second threshold, so as to approximate the terrain geometry.

3. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 2, characterized in that, Step 1.6: If a grid cell of the uniform Cartesian mesh intersects with a triangular facet in the terrain surface mesh or the distance is less than a first threshold, then the grid cells of the uniform Cartesian mesh are densified until the size of the grid cells of the uniform Cartesian mesh is less than a second threshold, in order to approximate the terrain geometry, wherein: The first threshold is 10m, and the second threshold is 10~20m; The encryption rule for the uniform Cartesian grid is to divide the grid cell of the uniform Cartesian grid in half along the x, y, and z directions respectively. If the encrypted grid cell does not meet the condition that the grid size is less than the second threshold, the grid cell is further divided until the size of the grid cell is less than the second threshold. Where x represents the horizontal east-west coordinate of the grid cell, y represents the horizontal north-south coordinate of the grid cell, and z represents the vertical height coordinate of the grid cell.

4. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 1, characterized in that, Step 2: Constructing the geometric model of the wind turbine blade, discretizing the geometric model of the wind turbine blade into a wind turbine blade mesh, generating the surface mesh of the wind turbine blade, filling the surface mesh to generate a volume mesh, and generating an anisotropic structured boundary layer mesh on the surface of the wind turbine blade. The surface mesh, volume mesh, and boundary layer mesh of the wind turbine blade constitute the structured or unstructured hybrid mesh of the wind turbine region, including the following steps: Step 2.1: Construct the initial geometric model of the wind turbine blade. Perform geometric cleaning and parameterization on the initial geometric model of the wind turbine blade. Use a non-uniform rational B-spline surface to describe the wind turbine blade with high precision to obtain the geometric model of the wind turbine blade. Step 2.2: Based on the geometric model of the wind turbine blade, generate a two-dimensional wind turbine blade surface mesh. The mesh unit is a quadrilateral or triangular structure. Fill the wind turbine blade surface mesh to generate the wind turbine blade body mesh, and generate a boundary layer mesh on the wind turbine blade surface.

5. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 4, characterized in that, Step 2.2: Based on the geometric model of the wind turbine blade, a two-dimensional wind turbine blade surface mesh is generated, with the mesh cells being quadrilateral or triangular structures. The wind turbine blade surface mesh is filled to generate the wind turbine blade body mesh, and a boundary layer mesh is generated on the wind turbine blade surface. This includes the following steps: Step 2.2.1: Generate an anisotropic structured blade surface boundary layer mesh on the surface of the wind turbine blade, with the mesh lines forming an angle ≥85° with the normal direction of the wind turbine blade surface; Step 2.2.2: Generate a boundary layer mesh of triangular prism shape at the connection points of each component of the wind turbine blade; Step 2.2.3: Adjust the fineness of the blade surface boundary layer mesh and the connection position boundary layer mesh in steps 2.2.1 and 2.2.2, according to the preset... The thickness of the first boundary layer mesh is adjusted, and the number of boundary layer meshes on the blade surface and at connection points is set to ≥30 layers, with a thickness gradient ratio ≤1.2, thus forming the boundary layer mesh on the wind turbine blade surface. This is used to measure the ratio of the distance from the center of the first grid to the surface of the fan to the viscous length of the airflow; Step 2.2.4: Generate a tetrahedral mesh, which smoothly transitions from the outer edge of the boundary layer mesh on the surface of the wind turbine blade to the far field, with a transition range of 10 times the blade chord length.

6. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 1, characterized in that, Step 3: Generate an anisotropic wake region O-type structure mesh. This involves constructing a wake volume by geometrically abstracting and extracting feature lines from the wind turbine wake region, transforming the complex wind turbine wake flow field into a geometric model. Based on this wake volume, construct O-type topological blocks. The topological structure of these O-type topological blocks is a closed O-type mesh in cross-section. Refine the O-type mesh to obtain the wake region O-type structure mesh, including the following steps: Step 3.1: Geometrically abstract the wind turbine wake region, representing it as a cylinder starting from a certain distance upstream of the impeller and extending downstream. Extract feature lines from the wind turbine wake region, extracting the wake centerline and boundary shear layer to construct the wake volume. Step 3.2: Based on the wake, construct an O-type topology block with the wind turbine's rotation axis as the central axis. The topology of the O-type topology block is a closed O-type mesh in cross-section. The diameter of the cylindrical computational domain of the O-type mesh is 10 times the diameter of the wind turbine rotor, and the length is 150 times the diameter of the wind turbine rotor. The distance from the rotational plane where the wind turbine is located to the inlet of the cylindrical computational domain is 60 times the diameter of the wind turbine rotor, and the distance from the outlet of the cylindrical computational domain is 90 times the diameter of the wind turbine rotor. Step 3.3: Dense the O-type mesh radially within a range of 1.2 times the hub radius centered on the hub and within a range of 0.8 to 1.2 times the impeller radius, and axially denser the O-type mesh within a range of 5 times the impeller diameter upstream to 90 times the impeller diameter downstream of the rotation plane where the wind turbine is located.

7. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 6, characterized in that, Step 3.3: Radially densifying the O-type mesh within a range of 1.2 times the hub radius and 0.8 to 1.2 times the impeller radius centered on the hub, and axially densifying the O-type mesh within a range of 5 times the impeller diameter upstream to 90 times the impeller diameter downstream of the wind turbine's rotation plane, includes the following steps: Step 3.3.1: Within a range of 1.2 times the hub radius centered on the hub, the O-type mesh is densified using a finer mesh of 0.2 times the hub diameter; within a range of 0.8 times the impeller radius to 1.2 times the impeller radius, the O-type mesh is densified using an ultrafine mesh of 0.1 times the blade tip chord length. Step 3.3.2: Based on the flow velocity attenuation rate and vortex structure dissipation scale, the O-type grid is non-uniformly segmented and refined with the rotation plane where the wind turbine is located as the 0 point. The area from 5D upstream to 5D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.01D. The area from 5D to 20D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.02D to 0.05D. The area from 20D to 90D downstream of the rotation plane where the wind turbine is located is refined with an axial step size of 0.1D. Where D is the diameter of the wind turbine rotor.

8. The multi-grid fusion method for numerical simulation of wind turbine wake in complex terrain as described in claim 1 or 6, characterized in that, The O-type topology block can also be a C-type or HO hybrid topology block.

Images