Offshore wind power plant micro-siting strategy based on improved wake flow model

Abstract

The invention relates to an offshore wind power plant micro-siting strategy based on an improved wake flow model, and the strategy comprises the steps: collecting, statistically analyzing the long-term wind speed and wind direction data of a target region, and constructing a corresponding probability density function model; based on a Jensen wake flow model and Gaussian distribution, an improved Jensen wake flow model and an improved Gaussian wake flow model are provided, a mathematical model of the wake flow effect of the wind turbine is established, the influence of wake flow on the output power of the wind turbine is quantified, and mutual interference between units is simulated; the wake flow modeling precision is improved, the turbulence intensity is introduced, the influence of the turbulence intensity is quantified, and a Reynolds average Navier-Stokes equation model is adopted to more accurately describe the wake flow effect between the wind turbine generators; and the probability density function model of the wind speed and the wind direction and the wake flow model are integrated, a wind turbine power curve is calculated, the total output power of the wind power plant is obtained, and the overall optimal wind turbine layout is obtained. The method can provide theoretical support and an optimization strategy for micro-siting of the wind power plant.

Description

Technical Field

[0001] The invention relates to the field of offshore wind farm micro-site selection, and in particular to a strategy for offshore wind farm micro-site selection based on an improved wake model. Background Art

[0002] As global energy demand continues to grow, wind energy, as a clean and renewable energy source, has gradually become an important source of electricity for various countries. In particular, offshore wind farms have attracted widespread attention due to their advantages such as higher wind speeds, more stable airflows, and less impact on land resources. Compared with onshore wind farms, offshore wind farms can utilize the vast space of the sea surface to achieve larger-scale wind energy utilization and reduce noise pollution to residential areas. However, the site selection and layout optimization of offshore wind farms still face many technical challenges, which have an important impact on power generation efficiency, investment costs, and operation and maintenance difficulties.

[0003] In the layout design of wind farms, the wake effect between wind turbines is a key factor that cannot be ignored. The wake effect refers to the formation of a "wake zone" downstream where the airflow velocity decreases and the turbulence intensity increases when wind turbines capture wind energy for power generation, thereby affecting the wind speed and power generation of downstream turbines. The wake effect will cause power loss in downstream wind turbines and increase the load fluctuation of wind turbines, ultimately affecting the overall power generation efficiency of the wind farm. Therefore, accurately simulating and quantifying the wake effect and optimizing the relative layout of wind turbines are one of the core challenges in wind farm design.

[0004] Traditional wake models, such as the Jensen wake model, are widely used in the analysis of the wake effect of wind farms. The Jensen model is a simple and effective wake model that is based on empirical formulas and can quickly calculate the wind speed attenuation in the wake area. However, the Jensen model has certain limitations in describing the wake expansion and turbulence characteristics, especially in offshore wind farms. Under the influence of complex meteorological conditions, traditional models often find it difficult to accurately capture the actual characteristics of the wake effect. At the same time, the Gaussian distribution wake model has gradually been applied. It is based on the turbulent diffusion theory and more accurately describes the lateral diffusion of the wake and the change of turbulence intensity, but the computational complexity is relatively high. Therefore, how to find a balance between accuracy and computational efficiency is an important direction in the study of wake models.

[0005] In addition, the spatial and temporal variations of wind speed and direction will also have a significant impact on the power generation efficiency of wind farms. Traditional site selection strategies usually rely on long-term wind speed and wind direction statistics in the target area, and describe the wind conditions in the area through probability density function models such as Weibull distribution. However, these methods fail to fully consider the turbulence intensity and complex wake interference effects, so there are certain errors in practical applications. Especially in the scenario of multi-unit collaborative power generation, ignoring the mutual influence of wake and turbulence may lead to poor site selection and layout optimization.

[0006] In order to solve the above problems, in recent years, some scholars and research institutions have begun to focus on the research work of improving the wake model. For example, by introducing turbulence intensity parameters to quantify the power effect of the wake on the wind turbine, and describing the mutual interference between turbines based on the Reynolds-averaged Navier-Stokes equations (RANS) model, the accuracy of wake modeling is improved. At the same time, some studies have also tried to combine the improved wake model with the probability density model of wind speed and wind direction to more accurately predict the power output of wind turbines. This method can not only effectively improve the power generation efficiency of wind farms, but also provide theoretical support for the micro-site selection of offshore wind farms. Summary of the invention

[0007] In response to the above problems, the present invention proposes a micro-site selection strategy for offshore wind farms based on an improved wake model. The long-term wind speed and wind direction data of the target area are collected and statistically analyzed to construct the corresponding probability density function model; based on the Jensen wake model and Gaussian distribution, an improved Jensen wake model and a Gaussian wake model are proposed to establish a mathematical model of the wind turbine wake effect, quantify the impact of the wake on the wind turbine output power, and simulate the mutual interference between units; improve the accuracy of wake modeling, introduce turbulence intensity, quantify its impact, and use the Reynolds-averaged Navier-Stokes equations (RANS) model to more accurately describe the wake effect between wind turbines; combine the probability density function model of wind speed and wind direction with the wake model to calculate the wind turbine power curve, and then obtain the total output power of the wind farm.

[0008] This method first collects and statistically analyzes the long-term wind speed and wind direction data of the target area, and constructs the corresponding probability density function model; based on the Jensen wake model and Gaussian distribution, an improved Jensen wake model and Gaussian wake model are proposed, and a mathematical model of the wind turbine wake effect is established to quantify the impact of the wake on the output power of the wind turbine and simulate the mutual interference between units; the accuracy of wake modeling is improved, turbulence intensity is introduced, its impact is quantified, and the Reynolds-averaged Navier-Stokes equations (RANS) model is used to more accurately describe the wake effect between wind turbines; the probability density function model of wind speed and wind direction and the wake model are combined to calculate the wind turbine power curve, and then the total output power of the wind farm is obtained, and the global optimal wind turbine layout is obtained.

[0009] The present invention is achieved through the following technical solutions:

[0010] A micro-site selection strategy for offshore wind farms based on an improved wake model includes the following steps:

[0011] step:

[0012] Step 1: Collect historical wind speed and wind direction data of the target area, perform statistical analysis on them, and construct a probability density function model pdf(v,θ) of wind speed v and wind direction θ;

[0013] Step 2: Based on the Jensen wake model and Gaussian distribution, an improved Jensen wake model and Gaussian wake model are proposed to establish a mathematical model of the wind turbine wake effect, quantify the impact of the wake on the wind turbine output power, and simulate the wind speed loss v caused by mutual interference between wind turbines def , and calculate the final wind speed v i ;

[0014] Step 3: Improve the wake modeling accuracy, introduce turbulence intensity TI(i), quantify its impact, and use the Reynolds-averaged Navier-Stokes equation model to more accurately describe the wake effect between wind turbines;

[0015] Step 4: Combine the probability density function model of wind speed and wind direction with the wake model to calculate the power curve P of the wind turbine wt (v i ), and the total output power of the wind farm P is obtained wf .

[0016] Furthermore, the probability density function model of wind speed and wind direction is a Weibull distribution model obtained by statistical analysis based on historical wind speed and wind direction data of the target area.

[0017]

[0018] Among them, k(θ) and c(θ) are the scale parameter and shape parameter respectively, both of which are functions of wind direction θ and can be estimated by maximum likelihood estimation of historical wind speed and direction information data:

[0019]

[0020] Among them, v θ,i is the i-th wind speed observed in direction θ, and n(θ) is the total number of wind speed data in direction θ.

[0021] Furthermore, in step 2: the Jensen wake model is improved to calculate the wake velocity of the wind turbine, and a mathematical model of the wind turbine wake effect is established:

[0022]

[0023] Among them, C t is the wind turbine output coefficient, k is the wake diffusion constant, which is taken as 0.05, d is the distance from the upstream wind turbine along the wind direction, r is the impeller radius, and A overlap is the affected area, calculated as follows:

[0024]

[0025] Where R is the wake radius, σ R is the wake angle, σ r is the blade angle, calculated as:

[0026]

[0027]

[0028] Among them, X refers to the overlapping distance between R and r;

[0029] Finally, the wind speed loss of the i-th wind turbine affected by all the remaining j (j≠i) wind turbines is calculated and the overall wind speed loss of all wind turbines is obtained:

[0030]

[0031] Improved Jensen wake model, the initial wind speed v of the i-th wind turbine in the wind farm 0 The wind speed v received after the wake loss i :

[0032] v i =v 0 -v de f.

[0033] Furthermore, in step 2: the improved Gaussian wake model is:

[0034]

[0035] Furthermore, in step 3: for the Reynolds stress term, the equation is further solved by introducing the w-∈ turbulence model:

[0036]

[0037] Among them, v i is the instantaneous velocity; t is the instantaneous velocity; is the time-averaged velocity; v i ′ is the fluctuating component of velocity; ρ is the gas density; is the time-averaged pressure; ν is the kinematic viscosity of the fluid; is the Reynolds stress term, which represents the additional stress caused by velocity fluctuations; i and j are index symbols used to represent the components of physical quantities in different directions; w is the turbulent kinetic energy; ∈ is the turbulent dissipation rate; C ∈1 and C ∈2 is an empirical constant;

[0038] P w is the generation term of turbulent kinetic energy, indicating the energy source of turbulence, and is obtained from Reynolds stress and velocity gradient:

[0039]

[0040] S w and S ∈ is an adaptive term and is calculated as follows:

[0041]

[0042] Among them, σ w is the Prandtl number of the turbulent kinetic energy; σ ∈ is the Prandtl number for the turbulent dissipation rate;

[0043]

[0044] Among them, C and p are empirical coefficients;

[0045]

[0046] Among them, σ v is the standard deviation of wind speed, indicating the amplitude of speed fluctuations.

[0047] Furthermore, the power curve of the wind turbine is simplified to a piecewise linear function:

[0048]

[0049] Among them, v in 、v N 、v out are the cut-in wind speed, rated wind speed, and cut-out wind speed of the wind turbine, respectively. wt,N is the rated power of the wind turbine;

[0050] For a wind farm with N wind turbines, the total output power P wf :

[0051]

[0052] Among them, f k (θ) is the frequency of occurrence of the kth wind direction; D is the total number of equal-width intervals into which the wind direction is discretized.

[0053] The beneficial effects of the present invention are as follows:

[0054] The present invention aims to provide a micro-site selection strategy for offshore wind farms based on an improved wake model, taking into account the probability density function and turbulence of wind speed and wind direction, using multiple wake models, optimizing the layout of wind turbines, and reducing the impact of the wake effect on the total power generation efficiency of the wind farm. Through this strategy, the layout of the wind farm can be optimized under different wind speed and wind direction conditions to maximize the power generation efficiency, while reducing the mutual interference problem between wind turbines in the wind farm, overcoming the shortcomings of the prior art.

[0055] The probability density function model of wind speed and wind direction is a Weibull distribution model obtained through statistical analysis based on historical wind speed and wind direction data of the target area.

[0056] The wake model is based on the existing wake model and proposes a new improved model for quantifying the influence of the wake effect on the output power of the wind turbine and simulating the mutual interference between units.

[0057] The application of fluid mechanics and turbulence models improves the accuracy of the wake model.

[0058] By combining the probability model of wind speed and wind direction with the wake model, the power curve of the wind turbine is finally obtained, and the output power of the entire wind farm is calculated to obtain the global optimal wind turbine layout. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0060] Figure 1It is an overall flow chart of a strategy for micro-site selection of offshore wind farms based on an improved wake model of the present invention;

[0061] Figure 2 Schematic diagram of Jensen wake model;

[0062] Figure 3 It is a schematic diagram of Gaussian wake model;

[0063] Figure 4 Schematic diagram of the RANS model. DETAILED DESCRIPTION

[0064] In order to make the above-mentioned purposes, features and advantages of the present invention more clearly understood, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings. The specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0065] like Figure 1 As shown in FIG. 1 , a micro-site selection strategy for offshore wind farms based on an improved wake model is applicable to micro-site selection for offshore wind farms, and specifically includes the following steps:

[0066] Step 1: Collect long-term wind speed and direction data in the target area, perform statistical analysis on them, build a probability density function model of wind speed and wind direction, and use the probability density function (pdf) of the Weibull distribution to describe the uncertainty of wind speed:

[0067]

[0068] Among them, k(θ) and c(θ) are the scale parameter and shape parameter respectively, both of which are functions of wind direction θ and can be estimated by maximum likelihood estimation of historical wind speed and direction information data:

[0069]

[0070]

[0071] Among them, v θ,i is the i-th wind speed observed in direction θ, n(θ) is the total number of wind speed data in direction θ. In order to facilitate calculation and subsequent iteration of the differential evolution algorithm using MATLAB, θ is divided into 0°, 10°, 20°, …, 340°, and 350° at intervals of 10°.

[0072] Step 2: Wake losses are an important factor in wind farm layout design. When two wind turbines are placed too close together, one behind the other, in the prevailing wind direction, the total amount of electricity produced is less than the sum of the electricity originally expected from each wind turbine operating individually, because the wind energy in the airflow available to the downwind wind turbine is reduced by the wind energy extracted by the upwind wind turbine. Typically, when a uniform incoming wind encounters a wind turbine, a linearly expanding wake appears behind the wind turbine.

[0073] The Jensen model is a one-dimensional linear wake model based on three assumptions: the initial wake diameter is equal to the rotor diameter; the wake boundary grows linearly; and the velocity distribution in the wake region is uniform. The wind speed in the wake region can be simply described by the following formula:

[0074]

[0075] Due to the limitations of the Jensen model and considering the possible angle between the wind turbine and the upstream wind direction, the Jensen wake model is improved to calculate the wake velocity of the wind turbine and establish a mathematical model of the wind turbine wake effect:

[0076]

[0077] Among them, C t is the output coefficient of the wind turbine, k is the wake diffusion constant, which is usually related to the atmospheric turbulence intensity. The commonly used empirical values ​​are 0.075 (for onshore wind turbines) or 0.04 to 0.05 (for offshore wind turbines). Since this paper is a micro-site selection for offshore wind farms, 0.05 is taken here. d is the distance from the upstream wind turbine along the wind direction, r is the impeller radius, and A overlap is the affected area, calculated as follows:

[0078]

[0079] Where R is the wake radius, σ R is the wake angle, σ r is the blade angle, calculated as:

[0080]

[0081] Here, X refers to the overlapping distance between R and r.

[0082] Finally, the wind speed loss of the i-th wind turbine affected by all the remaining j (j≠i) wind turbines is calculated and the overall wind speed loss of all wind turbines is obtained:

[0083]

[0084] Improved Jensen wake model, the initial wind speed v of the i-th wind turbine in the wind farm 0 The wind speed v received after the wake loss i :

[0085] v i =v 0 -v de f

[0086] A simplified schematic diagram of the Jensen wake model is shown below: Figure 2 .

[0087] In order to balance calculation accuracy and efficiency and adapt to different design stages and environmental conditions, the combination of multiple models can better describe the mutual interference between wind turbine wakes and adapt to complex terrain, thereby optimizing the overall performance of the wind farm.

[0088] The Gaussian wake model assumes that the wind speed of the wake is Gaussian distributed. Gaussian distribution is a widely existing distribution type in real life, but the real wake profile is not uniformly distributed. The wind speed at the edge of the wake gradually approaches the ambient wind speed, and the loss at the center of the wake is greater. It is suitable for describing the gradual distribution of wind speed on the wake cross section. The wind speed attenuation formula of the Gaussian wake model is:

[0089]

[0090] Where σ is the standard deviation of the wake, which represents the expansion of the wake radius R with distance d, usually expressed as:

[0091] σ=kd+R

[0092] According to the rule of thumb, about 68% of the values ​​in a Gaussian distribution are within one standard deviation σ around the mean, 95% are within two standard deviations, and 99.7% are within three standard deviations. In general, the model uses 2.58 times the standard deviation to represent the outer boundary of the wake:

[0093] 2.58σ=R

[0094] As the distance between the two wind turbines increases, the Gaussian distribution of the wake will become flatter and the loss of the central velocity will be smaller. Therefore, in order to further ensure high accuracy, the model uses an adaptive standard deviation to represent the outer boundary of the wake:

[0095] (3-e -(d-200) )σ=R

[0096] In this way, the improved Gaussian wake model is:

[0097]

[0098] A simplified schematic diagram of the Gaussian wake model is shown below: Figure 3 .

[0099] Step 3: Wind turbine wakes and turbulence are critical to understanding the performance of wind turbines and their impact on the surrounding environment. Wind turbine wakes refer to the area of ​​low-speed airflow behind the wind turbine due to kinetic energy conversion, while turbulence refers to the complex, irregular motion that is common in fluid flow, that is, the operation of wind turbine blades causes the airflow to separate, forming vortices and turbulence. The wake area is usually accompanied by high turbulence intensity because the airflow speed changes rapidly during the process of wind turbine extraction of wind energy, causing irregular vortices and momentum transfer.

[0100] Turbulence intensity (TI) is an important parameter describing turbulent fluctuations in the wake, which has a significant impact on the operation of downstream wind turbines. The turbulence intensity of the i-th wind turbine is defined as:

[0101]

[0102] Among them, σ v is the standard deviation of wind speed, indicating the amplitude of speed fluctuations.

[0103] The turbulence intensity in the wake usually decreases with increasing distance. This change can be described by an empirical formula. The common formula for the turbulence intensity at a distance d from the i-th wind turbine is:

[0104]

[0105] Among them, C and p are empirical coefficients, usually determined based on experimental data.

[0106] In order to describe the velocity and turbulence characteristics of the wake, the Reynolds average equation (RANS) is usually used. The core idea of ​​the RANS equation is to decompose physical quantities such as fluid velocity and pressure into average quantities and fluctuation quantities:

[0107]

[0108] Among them, v i is the instantaneous velocity; is the time-averaged velocity; v i ′ is the fluctuation component of velocity.

[0109] On this basis, the Reynolds average equation is obtained:

[0110]

[0111] Where ρ is the gas density; is the time-averaged pressure; ν is the kinematic viscosity of the fluid; is the Reynolds stress term, which represents the additional stress caused by velocity fluctuations. In the RANS equations, i and j are index symbols used to represent the components of physical quantities in different directions, which help to describe the three-dimensional flow of the fluid and the interaction of turbulence in different directions.

[0112] The Reynolds stress term represents the additional stress introduced by turbulence and has an important influence on the momentum transfer of turbulence. Since the Reynolds stress term introduces additional unknowns, it needs to be closed through a turbulence model.

[0113] The w-∈ turbulence model is a common turbulence closed model that describes the behavior of turbulence by introducing turbulent kinetic energy w and turbulent dissipation rate ∈. It assumes that the turbulent viscosity coefficient ν t is associated with w and ∈ to solve for the Reynolds stress term. Where:

[0114] The turbulent kinetic energy w represents the average energy of turbulent velocity fluctuations per unit mass of fluid and is defined as:

[0115]

[0116] The turbulence dissipation rate ∈ represents the speed at which turbulent kinetic energy is gradually dissipated from large-scale to small-scale eddies and is defined as:

[0117]

[0118] Turbulent viscosity coefficient ν t :

[0119]

[0120] Among them, C μ is an empirical constant, generally 0.09.

[0121] Governing equations of the w-∈ turbulence model

[0122]

[0123] Among them, C ∈1 and C ∈2 is an empirical constant, usually 1.44 and 1.92; P w is the generation term of turbulent kinetic energy, which indicates the energy source of turbulence and can be obtained from Reynolds stress and velocity gradient:

[0124]

[0125] In the control equations of turbulent kinetic energy w and turbulent dissipation rate ∈, adaptive terms can be introduced to allow the model to dynamically adjust the turbulence characteristics under different wind speeds, turbulence intensity, wind farm layout and other conditions. In this way, the wake interference between wind turbines can be simulated more accurately. Therefore, an adaptive term S is added to the turbulent kinetic energy equation and the turbulent dissipation rate equation. w and S ∈ .

[0126]

[0127] S w and S ∈ The calculation method is as follows:

[0128]

[0129] Among them, σ w is the Prandtl number of turbulent kinetic energy, generally taken as 1.0; σ ∈ is the Prandtl number of the turbulent dissipation rate, which is generally taken as 1.3;

[0130] The combination of RANS equations and the w-∈ turbulence model enables the turbulence problems in the engineering field to be effectively solved, especially in the wake simulation of wind farms, providing an important tool for analyzing the mutual interference between wind turbines and optimizing wind farm efficiency.

[0131] The turbulence intensity wake profile using the RANS equation and the w-∈ turbulence model is shown in Figure 4 .

[0132] Step 4: The power curve of the wind turbine (WT) can be simplified as a piecewise linear function:

[0133]

[0134] Among them, v in 、v N 、v out are the cut-in wind speed, rated wind speed, and cut-out wind speed of the wind turbine, respectively. wt,N is the rated power of the wind turbine.

[0135] For a wind farm with N wind turbines, the total output power P wf :

[0136]

[0137] Among them, f k (θ) is the frequency of occurrence of the kth wind direction; D is the total number of equal-width intervals into which the wind direction is discretized. In order to facilitate calculation and subsequent algorithm iteration using MATLAB, each 10° interval is also used here, that is: D = 0, 1, 2, …, 35.

[0138] In order to achieve more efficient and accurate calculation of wake effects during micro-site selection, the Jensen wake model, Gaussian wake model and RANS model can be combined to flexibly select appropriate wake calculation methods according to specific needs. The Jensen wake model has a fast calculation speed and is suitable for preliminary estimation, but it is limited in accuracy. The Gaussian wake model can provide more detailed wind speed attenuation information by describing the lateral diffusion and turbulent changes of the wake, which is suitable for scenarios with higher requirements. The RANS model helps to capture the complex characteristics of the wake more accurately by decomposing the average flow and turbulent fluctuations, but the calculation amount is large. Considering the accuracy and calculation efficiency of the above models comprehensively, various models can be reasonably selected according to actual scenarios to provide accurate and efficient wake calculations for wind turbine site selection in complex environments. By combining these wake models into the site selection strategy through iterative optimization algorithms, the calculation efficiency can be significantly improved while maintaining accuracy, thus providing a reliable foundation for subsequent micro-site selection.

[0139] Those skilled in the art can understand that the above description is only a preferred example of the invention and is not intended to limit the invention. Although the invention is described in detail with reference to the above examples, those skilled in the art can still modify the technical solutions recorded in the above examples or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, etc. made within the spirit and principle of the invention shall be included in the protection scope of the invention.

Claims

1. A micro-site selection strategy for offshore wind farms based on an improved wake model, characterized in that: The method comprises the following steps: Step 1: Collect historical wind speed and wind direction data of the target area, perform statistical analysis on them, and construct a probability density function model pdf(v,θ) of wind speed v and wind direction θ; Step 2: Based on the Jensen wake model and Gaussian distribution, an improved Jensen wake model and Gaussian wake model are proposed to establish a mathematical model of the wind turbine wake effect, quantify the impact of the wake on the wind turbine output power, and simulate the wind speed loss v caused by mutual interference between wind turbines def , and calculate the final wind speed v i ; Step 3: Improve the wake modeling accuracy, introduce turbulence intensity TI(i), quantify its impact, and use the Reynolds-averaged Navier-Stokes equation model to more accurately describe the wake effect between wind turbines; Step 4: Combine the probability density function model of wind speed and wind direction with the wake model to calculate the power curve P of the wind turbine wt (v i ), and the total output power of the wind farm P is obtained wf .

2. A micro-site selection strategy for offshore wind farms based on an improved wake model as claimed in claim 1, characterized in that: The probability density function model of wind speed and wind direction is a Weibull distribution model obtained by statistical analysis based on the historical wind speed and wind direction data of the target area. Among them, k(θ) and c(θ) are the scale parameter and shape parameter respectively, both of which are functions of wind direction θ and are estimated by maximum likelihood estimation of historical wind speed and direction information data: Among them, v θ,i is the i-th wind speed observed in direction θ, and n(θ) is the total number of wind speed data in direction θ.

3. A micro-site selection strategy for offshore wind farms based on an improved wake model as claimed in claim 2, characterized in that: In step 2: the Jensen wake model is improved to calculate the wake velocity of the wind turbine and establish a mathematical model of the wind turbine wake effect: Among them, C t is the wind turbine output coefficient, k is the wake diffusion constant, which is taken as 0.05, d is the distance from the upstream wind turbine along the wind direction, r is the impeller radius, and A overlap is the affected area, calculated as follows: Where R is the wake radius, σ R is the wake angle, σ r is the blade angle, calculated as: Among them, X refers to the overlapping distance between R and r; Calculate the wind speed loss of the i-th wind turbine under the influence of all the remaining j (j≠i) wind turbines and obtain the overall wind speed loss of all wind turbines: Improved Jensen wake model, the initial wind speed v0 of the i-th wind turbine in the wind farm is affected by the wake loss and the wind speed v i : v i =v0-v de f.

4. A micro-site selection strategy for offshore wind farms based on an improved wake model as claimed in claim 3, characterized in that: In step 2: the improved Gaussian wake model is:

5. A micro-site selection strategy for offshore wind farms based on an improved wake model as claimed in claim 4, characterized in that: In step 3: For the Reynolds stress term, the equation is further solved by introducing the w-∈ turbulence model: Among them, v i is the instantaneous velocity; t is the instantaneous velocity; is the time-averaged velocity; v i ′ is the fluctuating component of velocity; ρ is the gas density; is the time-averaged pressure; ν is the kinematic viscosity of the fluid; is the Reynolds stress term, which represents the additional stress caused by velocity fluctuations; i and j are index symbols used to represent the components of physical quantities in different directions; w is the turbulent kinetic energy; ∈ is the turbulent dissipation rate; C ∈1 and C ∈2 is an empirical constant; P w is the generation term of turbulent kinetic energy, indicating the energy source of turbulence, and is obtained from Reynolds stress and velocity gradient: S w and S ∈ is an adaptive term and is calculated as follows: Among them, σ w is the Prandtl number of the turbulent kinetic energy; σ ∈ is the Prandtl number for the turbulent dissipation rate; Among them, C and p are empirical coefficients; Among them, σ v is the standard deviation of wind speed, indicating the amplitude of speed fluctuations.

6. A micro-site selection strategy for offshore wind farms based on an improved wake model as claimed in claim 5, characterized in that: The power curve of a wind turbine is simplified to a piecewise linear function: Among them, v in 、v N 、v out are the cut-in wind speed, rated wind speed, and cut-out wind speed of the wind turbine, respectively. wt,N is the rated power of the wind turbine; For a wind farm with N wind turbines, the total output power P wf : Among them, f k (θ) is the frequency of occurrence of the kth wind direction; D is the total number of equal-width intervals into which the wind direction is discretized.

Images