A semi-empirical wake model optimization method for wind turbine based on CFD numerical simulation

Abstract

The application discloses a wind turbine semi-empirical wake model optimization method based on CFD numerical simulation, and belongs to the technical field of wind power generation, and comprises the following steps: performing CFD numerical simulation on a wake field, obtaining wake velocity field distribution data, analyzing defects of a traditional Jensen wake model, quantifying influences of each defect on model accuracy, and constructing an improved Jensen wake model; a cosine function model is used to describe radial distribution of wind speed in a wake area to replace a uniform distribution assumption, a power function model is used to describe axial expansion law of a radius in the wake area to replace a linear expansion assumption, an empirical relationship between key parameters of the power function model and environmental turbulence intensity is established to adapt to different environmental turbulence intensity working conditions, and the verified improved Jensen wake model is used as an optimized wind turbine semi-empirical wake model. The method can comprehensively improve wake velocity distribution, a wake boundary and turbulence adaptability.

Description

Technical Field

[0001] This invention relates to a semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation, belonging to the field of wind power generation technology. Background Technology

[0002] The wake effect of wind turbine generators is a core factor restricting wind farm micro-site selection, generator layout optimization, accurate power generation estimation, and generator fatigue life assessment. Upstream generator wakes can lead to reduced wind speeds and increased turbulence intensity in downstream generators, directly causing power generation losses and structural fatigue damage. Therefore, the accuracy of the wake model directly determines the economic efficiency and safety of wind farm design and operation.

[0003] Currently, the Jensen semi-empirical wake model is widely used in the engineering field. Due to its simple structure and high computational efficiency, it has been integrated into mainstream wind resource management software. However, this model is based on experimental data from small wind turbines, and its inherent shortcomings are becoming increasingly apparent with the increasing size of wind turbines and the increasing complexity of wind farm environments. First, it assumes a radially uniform distribution of wake wind velocity, which is significantly inconsistent with the actual valley-shaped distribution characterized by a low axis and high sides. Second, it assumes a linear expansion of the wake radius, underestimating diffusion near the wake and overestimating diffusion far from the wake, resulting in poor adaptability. Third, it fails to quantify the impact of environmental turbulence intensity on wake recovery, leading to significant errors under high and low turbulence conditions.

[0004] Computational Fluid Dynamics (CFD) numerical simulations can solve for detailed wake flow fields with high accuracy, but they are computationally expensive and time-consuming, making them unsuitable for large-scale wind farm engineering designs. Using high-fidelity CFD data for semi-empirical model calibration is a mainstream approach that balances accuracy and efficiency. Existing improvement schemes often only correct the expansion coefficient or superimpose turbulence effects, failing to achieve systematic optimization from the perspectives of radial distribution, expansion patterns, and turbulent coupling. Therefore, there is still significant room for improvement in the model's versatility and accuracy. Summary of the Invention

[0005] The purpose of this invention is to provide a semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation, which can comprehensively improve wake velocity distribution, wake boundary, and turbulence adaptability.

[0006] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for optimizing a semi-empirical wake model of a wind turbine generator based on CFD numerical simulation, comprising: A three-dimensional model and computational domain of a wind turbine generator were established. The Reynolds time-averaged method was used in conjunction with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator and obtain wake velocity field distribution data under different environmental turbulence intensities. Based on wake velocity field distribution data analysis, the traditional Jensen wake model is found to have shortcomings in the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities. The impact of each defect on the accuracy of the traditional Jensen wake model is also quantified. An improved Jensen wake model is constructed based on the impact of various defects on the accuracy of the traditional Jensen wake model: A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The improved Jensen wake model was validated using field measurement data, and the validated improved Jensen wake model was used as the optimized semi-empirical wake model for wind turbine generators.

[0007] In conjunction with the first aspect, further computational fluid dynamics numerical simulations include: A hybrid grid strategy is adopted to divide the computational domain. The region containing the wind turbine is set as the rotating domain and uses an unstructured grid, while the region outside the rotating domain is set as the stationary domain and uses a structured grid. A multi-reference-frame model is used to handle wind turbine rotation; The inlet boundary condition of the computational domain is set to the logarithmic wind profile of the neutral atmospheric boundary layer, the ground boundary condition is the standard wall function, the outlet boundary condition is the pressure outlet, and the side and top boundary conditions are symmetrical boundaries.

[0008] Combining the first aspect, the cosine function model is further as follows: ; in, Indicates the downstream axial position of the wind turbine Radial position Wind speed, Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the downstream axial position of the wind turbine Radius of the wake region Indicates the inflow wind speed.

[0009] Combining the first aspect, the further power function model is as follows: ; in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter of the power function model.

[0010] In conjunction with the first aspect, the empirical relationship between the undetermined wake expansion coefficient and the intensity of environmental turbulence is further expressed as: ; in, This indicates the intensity of environmental turbulence.

[0011] Combining the first aspect, the further improved Jensen wake model is as follows: ; in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter in the power function model. Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the inflow wind speed. This indicates the thrust coefficient of the wind turbine. Represents the empirical coefficient. This indicates the radius of the wind turbine's rotor. Indicates the downstream axial position of the wind turbine Radial position Local wind speed.

[0012] In conjunction with the first aspect, further, the field measured data includes field wind measurement data, lidar wind measurement experimental data, and wind tunnel experimental data. The verification indicators include the radial distribution of wind speed in the wake region, the axial attenuation distribution of wind speed in the wake region, and the radius of the wake region.

[0013] Secondly, the present invention provides a semi-empirical wake model optimization system for wind turbine generators based on CFD numerical simulation, comprising: The computational fluid dynamics numerical simulation module is used to establish a three-dimensional model and computational domain of the wind turbine generator set. It adopts the Reynolds time-averaged method and combines it with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator set and obtain wake velocity field distribution data under different environmental turbulence intensities. The defect analysis module is used to analyze the defects of the traditional Jensen wake model based on wake velocity field distribution data, such as the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities, and to quantify the impact of each defect on the accuracy of the traditional Jensen wake model. The model optimization module is used to construct an improved Jensen wake model based on the impact of various defects on the accuracy of the traditional Jensen wake model. A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The model validation module is used to validate the improved Jensen wake model using field measurement data, and the validated improved Jensen wake model is used as the optimized semi-empirical wake model for wind turbine generators.

[0014] Thirdly, the present invention provides a computer device, comprising: Storage medium: used to store computer programs; Processor: Used to execute the computer program to implement the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation as described in the first aspect.

[0015] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation as described in the first aspect.

[0016] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation as described in the first aspect.

[0017] Compared with the prior art, the beneficial effects of the present invention are: The present invention provides a semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation. It optimizes the turbulence model through CFD numerical simulation, accurately analyzes the defects of the Jensen model, and constructs an improved model from three aspects: radial, axial, and turbulent coupling. The accuracy is significantly better than the original model as verified by measured data. It is suitable for wind turbines of different sizes and complex turbulent environments, and balances computational efficiency and engineering accuracy. It can be directly applied to engineering scenarios such as micro-site selection of wind farms and power generation prediction, and has strong practicality. Attached Figure Description

[0018] Figure 1 This is a flowchart of the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in this embodiment of the invention. Figure 2 This is a schematic diagram comparing the axial wind speeds of the Jensen model and numerical simulation provided in this embodiment of the invention. Figure 3 This is a schematic diagram of the cosine model of wind speed radial distribution provided in an embodiment of the present invention, wherein (a), (b), (c), and (d) correspond to the wind speed radial distribution at axial positions of 2d, 4d, 8d, and 16d, respectively, and d is the diameter of the wind turbine. Figure 4 This is a schematic diagram comparing the wake radius power model and the linear model provided in this embodiment of the invention; Figure 5 These are the axial wind speed attenuation curves under different environmental turbulence intensities provided in the embodiments of the present invention. Among them, (a), (b), (c), and (d) correspond to the axial wind speed attenuation curves under four working conditions with environmental turbulence intensities of 5%, 7%, 9%, and 11%, respectively. Detailed Implementation

[0019] The technical solution of the present invention will be further described in detail below with reference to specific embodiments.

[0020] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Unless otherwise specified, embodiments of the present invention and the technical features thereof can be combined with each other.

[0021] This invention provides a method for optimizing a semi-empirical wake model of a wind turbine generator based on CFD numerical simulation, comprising: A three-dimensional model and computational domain of a wind turbine generator were established. The Reynolds time-averaged method was used in conjunction with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator and obtain wake velocity field distribution data under different environmental turbulence intensities. Based on wake velocity field distribution data analysis, the traditional Jensen wake model is found to have shortcomings in the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities. The impact of each defect on the accuracy of the traditional Jensen wake model is also quantified. An improved Jensen wake model is constructed based on the impact of various defects on the accuracy of the traditional Jensen wake model: A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The improved Jensen wake model was validated using field measurement data, and the validated improved Jensen wake model was used as the optimized semi-empirical wake model for wind turbine generators.

[0022] The semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in this invention achieves a comprehensive improvement in wake velocity distribution, wake boundary, and turbulence adaptability, meeting the requirements for engineering applications.

[0023] Figure 1 This is a flowchart illustrating the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in this embodiment. This flowchart only shows the logical sequence of the method in this embodiment; however, different methods may be used without conflict. Figure 1 Complete the steps shown or described in the order indicated.

[0024] The CFD numerical simulation-based semi-empirical wake model optimization method for wind turbine generators provided in this embodiment can be applied to a terminal and can be executed by a CFD numerical simulation-based semi-empirical wake model optimization system for wind turbine generators. This system can be implemented by software and / or hardware and can be integrated into the terminal, such as any tablet computer or computer device with communication capabilities.

[0025] In one possible embodiment, the optimization method for the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation specifically includes the following steps: Step 1: Establish a three-dimensional model and computational domain of the wind turbine generator set. Using the Reynolds time-averaged method and combined with the Reynolds stress turbulence model, perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator set to obtain wake velocity field distribution data under different environmental turbulence intensities. In this embodiment, the Reynolds time-averaged method is used, combined with the Reynolds stress turbulence model, and compared with the standard k-ε model to determine the Reynolds stress turbulence model as the optimal turbulence model. Computational fluid dynamics numerical simulation is performed on the wake field of a single wind turbine to obtain wake velocity field distribution data under different environmental turbulence intensity conditions.

[0026] Specifically, the computational fluid dynamics numerical simulation includes: using a hybrid mesh strategy to divide the computational domain, wherein the region containing the wind turbine is set as the rotating domain and uses an unstructured mesh, and the region outside the rotating domain is set as the stationary domain and uses a structured mesh; using a multi-reference frame model to handle the wind turbine rotation; setting the inlet boundary condition of the computational domain as the logarithmic wind profile of the neutral atmospheric boundary layer, the ground boundary condition of the computational domain as the standard wall function, the outlet boundary condition of the computational domain as the pressure outlet, and the side and top boundary conditions of the computational domain as symmetrical boundaries.

[0027] Step 2: Based on the wake velocity field distribution data, analyze the shortcomings of the traditional Jensen wake model in terms of the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities, and quantify the impact of each defect on the accuracy of the traditional Jensen wake model. Step 3: Construct an improved Jensen wake model based on the impact of various defects on the accuracy of the traditional Jensen wake model; In this embodiment, the construction of an improved Jensen wake model based on the impact of various defects on the accuracy of the traditional Jensen wake model includes the following steps: Step 1: Use a cosine function model to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model; In this embodiment, a cosine function-type radial wind speed distribution model is established to derive the relationship between average wind speed and axial wind speed.

[0028] Specifically, the cosine function model is as follows: ; in, Indicates the downstream axial position of the wind turbine Radial position Wind speed, Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the downstream axial position of the wind turbine Radius of the wake region Indicates the inflow wind speed.

[0029] Step 2: Use a power function model to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model; In this embodiment, a power function type wake radius model is established, and a scaling factor is introduced to adapt to any wind turbine diameter.

[0030] Specifically, the power function model is as follows: ; in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter of the power function model.

[0031] Step 3: Establish empirical relationships between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities.

[0032] In this embodiment, the empirical relationship between the key parameters of the fitting model and the intensity of environmental turbulence is expressed.

[0033] Specifically, the empirical relationship between the undetermined wake expansion coefficient and the intensity of environmental turbulence is as follows: ; in, This indicates the intensity of environmental turbulence.

[0034] set up , , , , A total of 5 operating conditions were used to obtain the wake velocity field database.

[0035] The traditional Jensen wake model, based on the assumption of radial uniformity, overestimates the axial wind speed. Axial linear expansion leads to a slower near-wake recovery and a faster far-wake recovery, resulting in uncoupled turbulence. It has poor adaptability to different working conditions.

[0036] The improved Jensen wake model is as follows: ; in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter in the power function model. Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the inflow wind speed. This indicates the thrust coefficient of the wind turbine. Represents the empirical coefficient. This indicates the radius of the wind turbine's rotor. Indicates the downstream axial position of the wind turbine Radial position Local wind speed.

[0037] The improved Jensen wake model uses a cosine function to describe the radial valley distribution of wind speed in the wake region, replacing the original uniform assumption. It also replaces the original linear expansion assumption with a scalable power function curve, adapting to any rotor diameter. A linear relationship between key parameters and environmental turbulence intensity is established, enabling the model to adapt to different turbulent conditions. By integrating radial distribution, wake radius, and turbulence corrections, an improved model that can be directly applied in engineering is formed.

[0038] Step 4: Validate the improved Jensen wake model using field measurement data, and use the validated improved Jensen wake model as the optimized semi-empirical wake model for wind turbine generators.

[0039] Specifically, the field measurement data includes field wind measurement data, lidar wind measurement experimental data, and wind tunnel experimental data. The verification indicators include the radial distribution of wind speed in the wake region, the axial attenuation distribution of wind speed in the wake region, and the radius of the wake region.

[0040] Validation shows that the improved Jensen wake model is more accurate than the traditional Jensen wake model, and the improvement is more significant when the turbulence is stronger.

[0041] In one possible embodiment, such as Figure 1 As shown, the optimization method for the semi-empirical wake model of wind turbine generators based on CFD numerical simulation specifically includes the following steps: S1: CFD numerical simulation and turbulence model selection; This embodiment selects the Nibe-B wind turbine from the Nibe wind farm in Denmark as the research object. This model is a three-bladed horizontal axis structure with a concrete tower and glass fiber composite blades. The airfoil adopts the NACA44 series, and the main parameters are shown in Table 1. The wind farm is located in a plain area with grassland, which is a flat terrain that can be considered to meet the condition of horizontal uniform inflow. Mature measured data are available for use in verifying the simulation results. The overall simulation process strictly follows the CFD numerical simulation standard specifications.

[0042] Table 1: Main Parameters of Nibe-B Wind Turbine .

[0043] First, 3D modeling and mesh generation were carried out. ANSYS ICEMCFD software was used to complete the full-size geometric model of the Nibe-B wind turbine, including key components such as blades, nacelle, and tower, to avoid geometric simplification that could lead to simulation distortion. Then, a cuboid computational domain was constructed with dimensions of 960m long, 320m wide, and 205m high. The center of the turbine hub was set as the origin. The distance from the turbine to the upstream boundary was 4 times the diameter, the downstream boundary was 20 times the diameter, and the left, right, and top boundaries were all 4 times the diameter. The bottom surface was the hub height. This design ensured sufficient space for wake development while avoiding an excessively large computational domain that would increase redundant computational load.

[0044] The mesh generation adopts a hybrid mesh strategy, dividing the computational domain into a rotating domain near the wind turbine and an external stationary domain: the rotating domain uses an unstructured mesh; the stationary domain uses a structured mesh, with the mesh being denser in the near-field and gradually less dense in the far-field.

[0045] Numerical solutions were performed using ANSYS FLUENT software, employing the Reynolds Stress Model (RSM). A standard k-ε model was also used as a control group. The inflow wind velocity was set to 8.5 m / s, covering five environmental turbulence intensities: 5%, 7%, 9%, 11%, and 13%. The inflow boundary was defined using a user-defined function (UDF) program, which included logarithmic wind profiles, turbulent kinetic energy, and dissipation rate distributions. The rotating domain was simulated using a multiple reference frame (MRF) model, with a residual convergence criterion set to 1 × 10⁻⁶. -5 .

[0046] The simulation results were validated based on measured data from the Nibe wind farm. The radial wind speed distribution at axial positions of 2.5d, 6d, and 7.5d was compared between the RSM model and the standard k-ε model, where d is the rotor diameter. The RSM model can accurately capture the asymmetric mixing characteristics of the wake and can serve as the core data support for subsequent wake model optimization.

[0047] S2: Jensen model defect analysis; (1) Radial distribution defects: The Jensen model assumes that the wind speed is uniformly distributed within a circular area covered by the wake region on any plane downstream of the wind turbine, parallel to the rotor, and that the wind speed magnitude is independent of radial position. However, CFD simulation results show that the wind speed in the wake region exhibits a valley-shaped symmetrical distribution: the wind speed is lowest at the axis, gradually increases with increasing radial distance, and returns to the inflow wind speed at the wake boundary.

[0048] (2) Axial expansion defects: The Jensen model assumes a linear expansion of the wake radius, as shown in the formula: , This represents the initial radius of the wake at the wind turbine plane. The coefficient of wake expansion is represented by , but CFD simulations show that the wake expansion rate gradually decreases with increasing axial position. The Jensen model assumes uniform wind speed across the entire cross-section, calculating the cross-sectional average wind speed, not the minimum axial wind speed. Since the wake radius directly determines the wind speed obtained from mass conservation, errors in the wake expansion rate calculation are simultaneously transmitted to the wind speed calculation. The near-wake linear model expands too slowly, while the far-wake model expands too quickly, ultimately causing a systematic deviation in the model's calculated axial wind speed recovery rate: the near-wake wind speed recovery is calculated too slowly, while the far-wake wind speed recovery is calculated too quickly. A comparison of the axial wind speeds from the Jensen model and numerical simulations is provided. Figure 2 As shown.

[0049] (3) Turbulence adaptability defects: Jensen model wake expansion coefficient It is only related to surface roughness, not coupled with environmental turbulence intensity, and ignores the dominant role of turbulence in wake mixing. Therefore, it cannot reflect the regulatory mechanism of turbulence mixing on wake expansion. CFD multi-condition simulations confirm that the higher the turbulence intensity, the more intense the airflow mixing, and the faster the wake expansion rate and wind speed recovery rate. However, the Jensen model uses the same [condition / feature] for all turbulent conditions. Values, even when individually fitted to different values The values ​​also cannot take into account both near and far wakes, and the higher the turbulence, the more the deviation between the linear assumption and the actual non-uniform expansion law is amplified, and the calculation error of wind speed and wake radius increases sharply under high turbulence environment.

[0050] S3: Improved Jensen model construction; To address the three core shortcomings of the Jensen model, an improved model was constructed based on CFD simulation data, focusing on radial distribution, expansion patterns, and turbulent coupling.

[0051] (1) Radial distribution optimization: The cosine function model replaces the uniformity assumption and proposes a certain axial position The radial distribution model of wind speed at a given location is as follows: ; in, Radial position The function, , , All parameters are yet to be determined.

[0052] The model will approximate the valley-shaped wind speed distribution curve with a cosine function, such as... Figure 3 As shown. This cosine function should satisfy the following condition: Independent variable of cosine curve The range is .

[0053] This segment of the cosine curve is a curve within half a function period, that is, it satisfies: .

[0054] when At that time, that is, at the boundary of the wake region, Restore to inflow velocity That is, satisfying: .

[0055] The derivation of the Jensen model is based on the law of conservation of mass, therefore it can be approximated as the calculated mass. This is the average wind speed within the wake region on this plane, meaning their mass flow rates are equal. Based on this principle, we can ensure that the areas enclosed by the velocity profiles of the cosine curve and the Jensen model curve are equal, i.e., satisfying: .

[0056] Solving the system of equations simultaneously, we obtain the values ​​of each parameter as follows: .

[0057] Substituting the parameter values ​​into the radial wind speed distribution model at a certain axial position, we obtain the radial wind speed distribution model: .

[0058] Compared to the original Jensen model's assumption of a radially uniform distribution, the new radial distribution model more accurately reflects the actual situation. Along the wind turbine axis, the minimum wind speed calculated by this model is very close to the results of numerical simulations.

[0059] (2) Axial expansion optimization: Abandoning the linear expansion assumption, a power function wake radius model is constructed, which takes the form: ; in, , , All parameters are yet to be determined.

[0060] This model will replace the original linear function curve with a power function curve to describe the wake region boundary, such as... Figure 4 As shown. This power function should satisfy the following conditions: The independent variable of the power function curve The range is ; The initial radius of the wake was obtained from numerical simulation. That is, satisfying: .

[0061] Solve for the parameters The value is: ; in, ,index This is an empirical constant obtained based on numerical simulation results, with a value of 2.3.

[0062] Substituting the parameters into the power function wake radius model, we obtain the power function: ; Among them, the undetermined coefficients The wake boundary was determined to be 3.3 by fitting the numerical simulation results of the environmental turbulence intensity of 9% using MATLAB.

[0063] The power-law model formula for the wake radius is a nonlinear function, and the wake shape described by this model does not possess such scalability. If the formula is changed... The shape of the function will change. To make this model applicable to wind turbines of any size, the independent variable in the model formula also needs to be adjusted. With dependent variable Before, divide by the scaling factor. This is to achieve scalability of the model. For any size wind turbine to be studied, the scaling factor is... Defined as the ratio of the radius of the wind turbine to the radius of the wind turbine in this case, that is: .

[0064] The resulting power-law model for the wake radius, applicable to any size, is as follows: ; The exponent 2.3 was determined by CFD fitting. This model shows a rapid expansion rate near the wake and a rate decay in the far wake, consistent with actual wake diffusion patterns. A scaling factor was introduced. The improved model is adapted to wind turbines of different diameters and its prediction of axial wind speed is more accurate than that of the original model.

[0065] (3) Turbulent coupling optimization: finding the coefficients With environmental turbulence intensity The relationship is given, and the calculation is provided. The empirical formula.

[0066] First, the wake region boundary was fitted under four operating conditions with environmental turbulence intensities of 5%, 7%, 9%, and 11%. As a result, the fitted Substituting the values ​​into the expression of the improved Jensen model, we can calculate... And calculated as the axial wind speed. ,like Figure 5 As shown. Based on the fitting results for multiple working conditions, the coefficients are... The value of is related to the environmental turbulence intensity The corresponding relationship was plotted as a curve, ultimately yielding the result based on the environmental turbulence intensity. Calculate coefficients Empirical formula: .

[0067] Based on the fitting results for the four operating conditions above, the axial wind speeds of the improved Jensen model for each condition match the axial wind speeds obtained from the numerical simulations better than both the Jensen model wind speed and the Jensen model axial wind speed. This indicates that the coefficients fitted under different environmental turbulence intensities are effective. Indeed, this allows the improved Jensen model to match the numerical simulation results under different working conditions.

[0068] (4) Complete Improvement Model: Integrating radial distribution, wake radius, and turbulence correction, an improved Jensen wake model is formed: .

[0069] Equation (1) represents the downstream axial position of the wind turbine. Formula for calculating the radius of the wake region; Equation (2) represents the downstream axial position of the wind turbine. Formula for calculating the average wind speed in the wake region; Equation (3) represents the downstream axial position of the wind turbine. Radial position Formula for calculating wind speed.

[0070] S4: Model Validation; A lidar wind measurement experiment was selected to fully verify the accuracy of the improved Jensen model. Compared with the original Jensen model, the optimization effect was highlighted.

[0071] LiDAR wind measurement was conducted at a wind farm in Jiuquan City, Gansu Province, China. The wind farm is located between two mountains, with stable wind direction and minimal variation, predominantly easterly. The surrounding terrain is flat, so the influence of topographic flow can be considered negligible, and the wind speed within the farm is suitable for studying the wake model of the wind turbines. The turbines are arranged in a typical quincunx pattern, with a spacing of 9 times the rotor diameter in the prevailing wind direction and 5 times the rotor diameter perpendicular to the prevailing wind direction. The rotor diameter of the wind turbine is 82m.

[0072] LiDAR (LiDAR) measurement uses pulsed laser scanning. The beam reflects off particles in the air, and the Doppler shift effect is used to measure the speed of dust movement, i.e., wind speed. LiDAR measures wind speeds from 40m to 300m in altitude, with a data sampling frequency of 1s, and can measure wind speeds from 0m / s to 80m / s. LiDAR can perform scanning measurements in four modes: PPI mode (elevation angle determined, azimuth measurement range set); RHI mode (azimuth angle determined, elevation angle measurement range set); DBS mode (vertical height wind profile measurement); and LOS mode (specified azimuth measurement mode).

[0073] Downstream measurement data of the wind turbine was selected within a measurement period, with an inflow wind speed of 6 m / s, a corresponding thrust coefficient of 0.79, and a turbulence intensity of 8%. Radial wind speed distribution curves at axial positions of 2, 5, and 7 times the rotor diameter are presented and compared with the original Jensen model and the improved Jensen model for verification.

[0074] Verification showed that, based on a comparison of wind speed distributions at various locations, the improved Jensen model exhibits some bias in the near-wake region, underestimating the wind speed deficit in the wake. However, it closely approximates the actual situation in the far-wake region. Overall, the improved Jensen model demonstrates a significant improvement in accuracy compared to the original model.

[0075] This invention provides a semi-empirical wake model optimization system for wind turbine generators based on CFD numerical simulation, comprising: The computational fluid dynamics numerical simulation module is used to establish a three-dimensional model and computational domain of the wind turbine generator set. It adopts the Reynolds time-averaged method and combines it with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator set and obtain wake velocity field distribution data under different environmental turbulence intensities. The defect analysis module is used to analyze the defects of the traditional Jensen wake model based on wake velocity field distribution data, such as the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities, and to quantify the impact of each defect on the accuracy of the traditional Jensen wake model. The model optimization module is used to construct an improved Jensen wake model based on the impact of various defects on the accuracy of the traditional Jensen wake model. A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The model validation module is used to validate the improved Jensen wake model using field measurement data, and the validated improved Jensen wake model is used as the optimized semi-empirical wake model for wind turbine generators.

[0076] The wind turbine semi-empirical wake model optimization system based on CFD numerical simulation provided in this embodiment of the invention can execute the wind turbine semi-empirical wake model optimization method based on CFD numerical simulation provided in this embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0077] This invention provides a computer device, comprising: Storage medium: used to store computer programs; Processor: Used to execute computer programs to implement the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in the embodiments of the present invention.

[0078] This invention provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in this invention.

[0079] This invention provides a computer program product, including a computer program that, when executed by a processor, implements the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation provided in this invention.

[0080] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0081] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0082] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0083] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0084] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing a semi-empirical wake model of a wind turbine generator based on CFD numerical simulation, characterized in that, include: A three-dimensional model and computational domain of a wind turbine generator were established. The Reynolds time-averaged method was used in conjunction with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator and obtain wake velocity field distribution data under different environmental turbulence intensities. Based on wake velocity field distribution data analysis, the traditional Jensen wake model is found to have shortcomings in the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities. The impact of each defect on the accuracy of the traditional Jensen wake model is also quantified. An improved Jensen wake model is constructed based on the impact of various defects on the accuracy of the traditional Jensen wake model: A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The improved Jensen wake model was validated using field measurement data, and the validated improved Jensen wake model was used as the optimized semi-empirical wake model for wind turbine generators.

2. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 1, characterized in that, Computational fluid dynamics numerical simulations include: A hybrid grid strategy is adopted to divide the computational domain. The region containing the wind turbine is set as the rotating domain and uses an unstructured grid, while the region outside the rotating domain is set as the stationary domain and uses a structured grid. A multi-reference-frame model is used to handle wind turbine rotation; The inlet boundary condition of the computational domain is set to the logarithmic wind profile of the neutral atmospheric boundary layer, the ground boundary condition is the standard wall function, the outlet boundary condition is the pressure outlet, and the side and top boundary conditions are symmetrical boundaries.

3. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 1, characterized in that, The model for the cosine function is: ； in, Indicates the downstream axial position of the wind turbine Radial position Wind speed, Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the downstream axial position of the wind turbine Radius of the wake region Indicates the inflow wind speed.

4. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 1, characterized in that, The power function model is: ； in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter of the power function model.

5. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 4, characterized in that, The empirical relationship between the undetermined wake expansion coefficient and the intensity of environmental turbulence is as follows: ； in, This indicates the intensity of environmental turbulence.

6. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 1, characterized in that, The improved Jensen wake model is as follows: ； in, Indicates the downstream axial position of the wind turbine Radius of the wake region The scaling factor is defined as the ratio of the rotor radius of the target wind turbine to be optimized to the rotor radius of the standard wind turbine used to construct the power function model. The undetermined wake expansion coefficient, which is related to the intensity of environmental turbulence, is a key parameter in the power function model. Indicates the downstream axial position of the wind turbine Average wind speed in the wake region Indicates the inflow wind speed. This indicates the thrust coefficient of the wind turbine. Represents the empirical coefficient. This indicates the radius of the wind turbine's rotor. Indicates the downstream axial position of the wind turbine Radial position Local wind speed.

7. The method for optimizing the semi-empirical wake model of a wind turbine generator based on CFD numerical simulation according to claim 1, characterized in that, The field measured data includes field wind measurement data, lidar wind measurement experimental data, and wind tunnel experimental data. The verification indicators include the radial distribution of wind speed in the wake region, the axial attenuation distribution of wind speed in the wake region, and the radius of the wake region.

8. A semi-empirical wake model optimization system for wind turbine generators based on CFD numerical simulation, characterized in that, include: The computational fluid dynamics numerical simulation module is used to establish a three-dimensional model and computational domain of the wind turbine generator set. It adopts the Reynolds time-averaged method and combines it with the Reynolds stress turbulence model to perform computational fluid dynamics numerical simulation on the wake field of a single wind turbine generator set and obtain wake velocity field distribution data under different environmental turbulence intensities. The defect analysis module is used to analyze the defects of the traditional Jensen wake model based on wake velocity field distribution data, such as the radial distribution of wind speed in the wake region, the axial expansion law of the wake region radius, and the response to different environmental turbulence intensities, and to quantify the impact of each defect on the accuracy of the traditional Jensen wake model. The model optimization module is used to construct an improved Jensen wake model based on the impact of various defects on the accuracy of the traditional Jensen wake model. A cosine function model is used to describe the radial distribution of wind speed in the wake region, replacing the uniform distribution assumption in the traditional Jensen wake model. A power function model is used to describe the radial axial expansion law of the wake region, replacing the linear expansion assumption in the traditional Jensen wake model. Establish an empirical relationship between the key parameters of the power function model and the intensity of environmental turbulence, so that the improved Jensen wake model can adapt to different environmental turbulence intensities. The model validation module is used to validate the improved Jensen wake model using field measurement data, and the validated improved Jensen wake model is used as the optimized semi-empirical wake model for wind turbine generators.

9. A computer device, characterized in that, include: Storage medium: used to store computer programs; Processor: Used to execute the computer program to implement the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the semi-empirical wake model optimization method for wind turbine generators based on CFD numerical simulation as described in any one of claims 1 to 7.

Images