Wind turbine blade airfoil design method

Abstract

The invention discloses a wind turbine blade airfoil design method, and belongs to the technical field of wind turbine engineering. The wind turbine blade airfoil design method comprises the following steps that a sample point set of a blade airfoil is obtained, and aerodynamic parameters and structural parameters of the blade airfoil are calculated; according to the geometric parameters of the sample point set and the aerodynamic parameters and structural parameters of the blade airfoil, establishing a proxy model of the blade airfoil; and an optimization algorithm considering structural parameters is adopted to carry out optimization design on the agent model, and an optimization design airfoil profile is obtained. When the wind turbine blade airfoil is designed, the aerodynamic performance and the structural strength of the blade airfoil are comprehensively considered, the aerodynamic performance of the blade airfoil is improved, and the structural stability of the blade airfoil is enhanced, so that the blade airfoil with more excellent comprehensive performance is obtained, and the overall performance of the blade is enhanced.

Description

Technical Field

[0001] The invention belongs to the technical field of wind turbine engineering, and in particular relates to a method for designing a wind turbine blade airfoil. Background Art

[0002] Wind turbine blades are made up of different types of airfoils stacked along the span direction. Therefore, the aerodynamic performance of the two-dimensional cross-section airfoil determines the utilization rate of wind energy. Designing an airfoil with good aerodynamic performance is the key to enhancing the utilization rate of wind energy of the blade. Therefore, the current airfoil optimization is aimed at improving its aerodynamic performance. However, during the rotation of the blade, due to the uncertainty of wind direction, it will be subject to complex unsteady aerodynamic loads, which puts higher requirements on the structural performance of the blade. The structural performance of the blade directly determines the economy, safety and operation stability of the whole machine. In the wind farms currently in actual operation, the main part that causes the failure and shutdown of wind turbine generator sets is the blade. The cracking and breaking caused by the insufficient structural strength of the blade design often occur. The advantages and disadvantages of the blade structural performance are mainly reflected in the cross-sectional stiffness of the airfoil. The larger the cross-sectional stiffness, the stronger the ability to resist deformation when subjected to external loads. Therefore, the structural strength design of the cross-sectional airfoil of the wind turbine blade is very important, which plays a decisive role in improving the overall performance of the wind turbine blade and has a significant guiding significance for the sustainable development of the wind turbine industry. The existing inventions in the same field only consider the stability of the airfoil during aerodynamic design, but do not consider the structural performance of the airfoil, which has design limitations in practical applications. Summary of the invention

[0003] The present invention provides a wind turbine blade airfoil design method to solve the technical problem that in the prior art, only the aerodynamic performance of the blade airfoil is considered during the design, but the structural performance of the blade airfoil is not considered, resulting in low overall performance of the blade.

[0004] The present invention solves the above technical problems through the following technical solutions:

[0005] A method for designing a wind turbine blade airfoil is provided, comprising the following steps:

[0006] Acquiring a sample point set of a blade airfoil, and calculating aerodynamic parameters and structural parameters of the blade airfoil;

[0007] Establishing a proxy model of the blade airfoil according to the geometric parameters of the sample point set, the aerodynamic parameters of the blade airfoil and the structural parameters;

[0008] The proxy model is optimized by using an optimization algorithm that takes the structural parameters into consideration to obtain an optimized airfoil.

[0009] Preferably, the method of obtaining the sample point set of the blade airfoil comprises the steps of: performing geometric parameterization processing on the blade airfoil to obtain an optimization space; and extracting sample points in the optimization space to form the sample point set.

[0010] Preferably, the calculation of the aerodynamic parameters and structural parameters of the blade airfoil includes the steps of: establishing a grid model of the blade airfoil, and performing mechanical analysis on the grid model to obtain the aerodynamic parameters; obtaining geometric data of the cross section of the blade airfoil, and calculating the structural parameters.

[0011] Preferably, the aerodynamic parameters include lift coefficient, drag coefficient and lift-to-drag ratio; and / or, the structural parameters include section moment of inertia and torsional moment of inertia.

[0012] Preferably, according to the geometric parameters of the sample point set, the aerodynamic parameters of the blade airfoil and the structural parameters, a proxy model of the blade airfoil is established, comprising the steps of: obtaining an input vector corresponding to each sample point in the sample point set according to the geometric parameters; obtaining a response value of each sample point according to the aerodynamic parameters and the structural parameters; and establishing a Kriging model of the blade airfoil according to the input vector and the response value;

[0013] Preferably, the wind turbine blade airfoil design method also includes the steps of: calculating the EI function based on the Kriging model obtained based on the initial sample point set; calculating the maximum value of the EI function using the simplex method, and adding the sample point corresponding to the maximum value to the sample point set; repeating the above steps, continuously adding new sample points until the generated Kriging model meets the preset accuracy.

[0014] Preferably, the optimization algorithm taking into account the structural parameters is used to optimize the proxy model to obtain the optimized airfoil, including the steps of: establishing an optimization objective function, taking the maximum objective function as the optimization target, and obtaining the optimized airfoil, wherein the objective function is:

[0015]

[0016] In the formula, C l With C d are the lift coefficient and drag coefficient of the blade airfoil, α1 and α2 are the two selected angles of attack, I x with I y is the section inertia moment of the blade airfoil cross section around the x-axis and the y-axis, J is the torsional inertia moment of the blade airfoil cross section around the centroid, w1, w2, w3 and w4 are weight coefficients, and w1+w2+w3+w4=1.

[0017] Preferably, the step of establishing an optimization objective function, taking the maximum of the objective function as the optimization objective, and obtaining the optimized airfoil also includes the steps of: establishing an optimization mathematical model, taking the maximum of the optimization objective function as the optimization objective, and using a global optimization algorithm to optimize the blade airfoil, wherein the optimization mathematical model is:

[0018]

[0019] Where: t max With t cmax are the maximum relative thickness and the maximum thickness of the airfoil respectively, low and up represent the upper and lower limits of the value range, and the superscript 0 represents the parameter of the reference airfoil; set the constraints that the optimization target needs to meet, and the constraints are that the lift coefficient increases, the drag coefficient decreases, and the section moment of inertia and the torsional moment of inertia increase; judge whether the optimization target meets the constraints, if so, output the optimized airfoil, if not, continue to add sample points until the constraints are met.

[0020] Preferably, the wind turbine blade airfoil design method further comprises the step of: if the error between the values ​​output by two adjacent calculations of the mathematical optimization model is less than the convergence error, then outputting the optimization result and terminating the optimization process.

[0021] Preferably, the wind turbine blade airfoil design method also includes the steps of: parameterizing the airfoil blade using a CST parameterization method; and / or acquiring the sample point set using an optimized Latin hypercube sampling method; and / or establishing a grid topology model of the blade airfoil in ANSYS ICEM software, generating a structural grid, setting boundary conditions in FLUENT software, and performing flow field calculation and analysis.

[0022] The positive and progressive effects of the present invention are:

[0023] The present invention comprehensively considers the aerodynamic performance and structural strength of the blade airfoil when designing the wind turbine blade airfoil, which not only improves the aerodynamic performance of the blade airfoil, but also enhances the structural stability of the blade airfoil, thereby obtaining a blade airfoil with better comprehensive performance and enhancing the overall performance of the blade. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The figure is a flow chart of the wind turbine blade airfoil design method of the present invention.

[0025] Figure 2 Schematic diagram of the watershed grid division of the airfoil of the present invention.

[0026] Figure 3 It is a comparison diagram of the geometric shapes of the reference airfoil and the optimized airfoil of the present invention.

[0027] Figure 4 This is a comparison chart of the pressure coefficients of the reference airfoil and the optimized airfoil of the present invention at an angle of attack of 4°.

[0028] Figure 5 This is a comparison chart of the pressure coefficients of the reference airfoil and the optimized airfoil of the present invention at an angle of attack of 8°.

[0029] Figure 6 This is a schematic diagram of replacing the airfoil to the 68-100% position of the blade according to the present invention. DETAILED DESCRIPTION

[0030] The present invention will be described more clearly and completely below by way of embodiments in conjunction with the accompanying drawings, but the present invention is not limited to the scope of the embodiments.

[0031] Initial airfoil selection is the first step in the optimization process. Usually, basic airfoils are selected from a standard airfoil database (such as the NACA airfoil database) and preliminary screening is performed based on the operating conditions and design requirements of the wind turbine. This embodiment takes the NACA64618 airfoil as the research object, with a Reynolds number of Re = 6×10 6 , 4° and 8° attack angles are selected to perform double-attack angle aerodynamic design on the airfoil. The flow chart of the wind turbine blade airfoil design method is shown in Figure 1 shown.

[0032] Step (1): Parameterization of airfoil geometry. The airfoil geometry is described by parametric methods. Commonly used methods include CST method, polynomial representation, etc. Here, the CST parameterization method is used to parameterize the airfoil, and the upper and lower wing surface expressions are:

[0033] y=C(x)·S(x)+x·Δy te

[0034]

[0035] Where x, y are the coordinate points on the airfoil surface; C(x), S(x) are the class function and shape function; Δy te is the relative thickness of the trailing edge; the subscripts u and l represent the upper and lower wing surfaces of the airfoil respectively; N1 and N2 represent the types of geometric shapes; i is the polynomial index; n u and n l is the polynomial order; A u , A l are the shape function coefficients of the upper and lower surfaces respectively; is the Bernstein polynomial; is the number of combinations.

[0036] In this embodiment, a fifth-order CST parameterization method is used to fit the upper and lower airfoils, with a total of 12 design variables.

[0037] Step (2): Determine the optimization space. Set the design variable range of the airfoil geometry coefficient and use it as the optimization space. In this embodiment, the geometry coefficient design variable is the airfoil shape function coefficient A u With A l , the optimal space is A u With A l Specifically, the optimization space is a 20% change range of the design variables.

[0038] Step (3): Experimental design obtains an initial set of sample points. Determine the initial number of samples, which can be evenly distributed or selected by methods such as Latin hypercube sampling (LHS). In this embodiment, a number of sample points are selected in the airfoil design space, and experiments or numerical simulations are performed on these sample points to obtain response values. Subsequently, the sample points will be selected in several combinations within the value range of the design variables for simulation (such as CFD simulation) to obtain the corresponding aerodynamic performance parameters (such as lift coefficient and drag coefficient). Specifically, in the design optimization space, 100 groups of sample points are selected through the optimized Latin hypercube sampling method to form the initial sample point set. Compared with the original Latin hypercube sampling, the result has better spatial uniformity.

[0039] Step (4): Establishment of airfoil mesh model and solution of flow field parameters. Establish an airfoil mesh model, including the airfoil calculation domain and mesh division, and solve the flow field parameters and aerodynamic parameters. Specifically, mesh the geometric model to generate a finite element model; the accuracy of mesh division will affect the accuracy of the analysis and the calculation efficiency, so it is necessary to balance the mesh density and the calculation resources.

[0040] Commonly used software for meshing include: ANSYS, Abaqus, COMSOL Multiphysics. In this embodiment, a mesh topology model is established for the airfoil calculation domain in ANSYS ICEM software and a structural mesh is generated, and then boundary conditions are set in ANSYS FLUENT software and flow field calculation and analysis are performed.

[0041] In this embodiment, according to the design variables in the sample point set, a geometric model is established in the ANSYS ICEM software, and the flow field meshing results are shown as follows: Figure 2 After the meshing, the mesh file is imported into ANSYSFLUENT software to calculate the lift coefficient, drag coefficient and moment coefficient to obtain the aerodynamic parameters corresponding to all the sample points sampled in step (3).

[0042] Step (5): Calculation of airfoil structural parameters. The moment of inertia, which is positively correlated with the cross-sectional stiffness, is used as a structural performance index. As a thin-walled structure, the larger the cross-sectional moment of inertia and the torsional moment of inertia of the airfoil, the less likely it is to deform. In this embodiment, MATLAB software is used to calculate the cross-sectional moment of inertia and the torsional moment of inertia of the airfoil, and a program is written to represent the structural performance of different airfoils.

[0043] Specifically, in order to obtain the structural parameters corresponding to all sample points sampled in step (3), the section inertia moment and torsional inertia moment of each group of sample points airfoil around the x-axis and y-axis are calculated in MATLAB software. The calculation method is as follows:

[0044]

[0045] In the formula, S x With S y is the static distance around the x-axis and y-axis; c With y c I is the horizontal and vertical coordinates of the centroid; x with I y is the section inertia moment around the x-axis and y-axis; J is the torsional inertia moment around the centroid; s is the length of the airfoil section; t is the thickness of the thin-walled material of the airfoil section. It is assumed that the thickness is consistent everywhere inside during calculation.

[0046] There are two methods to calculate the area and circumference of the airfoil of the blade cross section: one is to solve the definite integral method, and the other is the successive trapezoidal segment summation method. The direct definite integral method is used to solve the calculation formulas for the half circumference and half area of ​​the airfoil as follows, where l 12 With s 12 are the semi-perimeter and semi-area respectively.

[0047]

[0048] Step (6): Establishment of Kriging surrogate model. Surrogate model is a simplified model used to replace complex and computationally expensive simulation models, and is intended to provide fast approximate predictions by learning the relationships in simulation data. Surrogate models are widely used in optimization design, especially in computational fluid dynamics (CFD) and structural optimization. Common surrogate models include response surface model (RSM), Kriging model (Kriging), radial basis function (RBF), support vector machine (SVM), etc. The surrogate model used in this embodiment is Kriging model (Kriging), also known as Gaussian Process Regression (GPR). The core idea of ​​Kriging surrogate model is to use spatial autocorrelation to predict the value of unknown points, which can not only provide predicted values, but also provide estimates of prediction uncertainty.

[0049] Specifically, based on the 100 sets of geometric parameter design variable sample points described in step (3), combined with the aerodynamic parameters and structural parameters corresponding to each set of sample points calculated in steps (4) and (5), a Kriging proxy model is established. The following is a specific description:

[0050] The output result of step (3) is used as the variable sample point; the output results of step (4) and step (5) are used as the aforementioned sample point response values; the output result of step (3) is 100 sets of geometric parameters in the initial sample point set obtained by sampling; the output result of step (4) is the calculated aerodynamic parameters, that is, the lift coefficient, drag coefficient and moment coefficient of the airfoil; the output result of step (5) is the calculated structural parameters, that is, the section moment of inertia and torsional moment of inertia of the airfoil.

[0051] Furthermore, MATLAB software was used to construct and calculate the Kriging model.

[0052] Steps to construct a Kriging surrogate model:

[0053] 1. Data preparation: Generate design variable sample points and collect response data (i.e., sample point response values);

[0054] 2. Select kernel function: Define the kernel function to capture the correlation structure in the data;

[0055] 3. Build the Kriging model: use sample point data for training;

[0056] 4. The sum of the weights is 1, and the minimum prediction variance is achieved by solving the Kriging equation.

[0057] In summary, the integrated test design in MATLAB software includes: meshing, aerodynamic and structural calculations, data processing and other processes; mainly using MATLAB software to call ICEM software script to divide the mesh, then calling Fluent software to solve the flow field data, and then reading the aerodynamic result file to complete the structural parameter S x ,S y ,I x ,I y ,J is calculated, and then the data is analyzed, and the samples are calculated in a continuous cycle.

[0058] Step (7): Adding point criteria to improve the accuracy of the proxy model. Through iteration, the adding point criteria is maximized in each step, the next evaluation point is found, and the proxy model is updated to improve the accuracy of the Kriging model. The principle is to calculate the maximum value of the expected improvement function of the proxy model.

[0059] In the training process of this embodiment, the EI point addition optimization criterion is used to improve the accuracy of the Kriging model. The EI point addition optimization can be briefly summarized as follows: after calculating the EI function on the Kriging model trained on the initial sample set, the simplex method is used to calculate the maximum value of the EI function, and the sample points corresponding to the maximum value are added to the sample set. The model is continuously trained, and new sample points are continuously added until a model that meets the accuracy is generated.

[0060] The predicted value y(X) of any point X in the Kriging model follows a Gaussian distribution, that is, Its probability density function is:

[0061]

[0062] If the minimum value of the objective function is y min , then the expected improvement I(X) is expressed as:

[0063] I(X)=y min -y(X)

[0064] Then the expected improvement I(X) is expressed as:

[0065]

[0066] Where: Φ and φ are the cumulative function and probability density function of the standard normal distribution respectively.

[0067] Step (8): Use linear weighting to establish the optimization target. Based on the high-confidence kriging proxy model established in steps (6) and (7), the airfoil is designed with multiple angles of attack for aerodynamic and structural design. In terms of aerodynamic performance, a multi-angle combination design is used. In terms of structural strength, the moment of inertia is added to the optimization target. After the combination, the optimization target is linearly weighted. The optimization target is:

[0068]

[0069] Among them, C l With C d are the lift coefficient and drag coefficient of the airfoil, α1 and α2 are the two selected angles of attack, I x with I y is the section moment of inertia around the x-axis and y-axis, J is the torsional moment of inertia around the centroid, w1, w2, w3 and w4 are weight coefficients; when weighting different targets, the weight coefficients must be regularized, w1, w2, w3 and w4∈[0,1], and w1+w2+w3+w4=1; different weight coefficients are assigned according to the applicable scope of the airfoil.

[0070] Step (9): Establish an optimization mathematical model. Taking the maximum objective function as the optimization goal, the constraints are that the lift coefficient increases, the drag coefficient decreases, and the moment of inertia and torsional moment of inertia increase, and a global optimization algorithm is used to optimize the airfoil design; in this embodiment, a genetic algorithm is used for optimization.

[0071] The genetic algorithm parameters were set as follows: initial population size 1000, evolution 200 generations, crossover probability 0.3, mutation probability 0.02, convergence error 1×10 -10 .

[0072] The optimization mathematical objectives and constraints are:

[0073] Object:Maximize F(x)=0.3×(C l / C d ) α＝4° +0.3×(C l / C d ) α＝8° +0.2(I x +I y )+0.2×J

[0074]

[0075] Where: t max With t cmax and are the positions of the maximum relative thickness and maximum thickness of the airfoil respectively, C is the chord length of the cross-section airfoil, and the superscript 0 represents the parameters of the reference airfoil.

[0076] Step (10): Continue to add 100 groups of sample points using the EI point addition criterion, set the convergence error, and determine whether the optimization target converges. If the condition is not met, continue to add sample points until the condition is met and the loop is exited to obtain the optimized airfoil. Step 10 includes the following sub-steps:

[0077] Step 10.1 sets the constraints that the optimization objective must meet to be convergence, and the convergence accuracy must be such that the error between two consecutive optimal values ​​is less than the convergence error or the maximum number of sample points is met;

[0078] Step 10.2: Determine whether the optimization target converges:

[0079] If not, return to step (9);

[0080] If yes, proceed to the next step 10.3;

[0081] Step 10.3: Determine whether the optimization target meets the convergence accuracy requirement:

[0082] If not, return to step (6);

[0083] If yes, exit the loop and output the optimization result.

[0084] Step (11): Replace the optimized airfoil with the original position of the blade reference airfoil, and use the ANSYS static solution module to calculate its flapping displacement, swing displacement and torsion angle.

[0085] Comparison of the geometric shapes of the reference airfoil NACA64618 and the optimized airfoil opt-180 Figure 3 As shown in the figure, the leading edge and upper surface of the optimized airfoil are fuller, and the thickness of the upper surface increases. The thickness of the baseline airfoil is 18.01%, and the maximum thickness is at 35%. The thickness of the optimized airfoil is 18.84%, and the maximum thickness is at 36.48%.

[0086] Figure 4 The pressure coefficient of the reference airfoil and the optimized airfoil at the design angle of attack is compared. The upper side of the curve in the figure is the suction surface, and the lower side is the pressure surface. The difference represents the pressure difference on the airfoil surface. The greater the leading edge pressure difference, the better the aerodynamic performance of the airfoil. As can be seen from the figure, the pressure coefficient of the pressure surface has not changed significantly, but the pressure coefficient of the suction surface has decreased significantly, thus increasing the pressure difference on both sides of the airfoil, thereby improving the aerodynamic performance.

[0087] Table 1 shows the aerodynamic performance comparison of the baseline airfoil and the optimized airfoil at the design angle of attack. It can be seen from the table that this optimization method improves the maximum lift coefficient and lift-to-drag ratio of the optimized airfoil at the two design angles of attack, while keeping the drag coefficient unchanged.

[0088] Table 1 Comparison of aerodynamic performance between baseline airfoil and optimized airfoil

[0089]

[0090] Replace the airfoil at 68-100% of the NREL 5MW wind turbine blade. Figure 6 As shown in the figure, the flapping displacement, swing displacement and tip torsion angle are calculated. The blade flapping direction is perpendicular to the wind rotor plane, and the swing direction is parallel to the wind rotor plane. Table 2 compares the displacement and torsion angle of the reference blade and the optimized blade. It can be seen from the data in the table that although the peak values ​​of the swing and flapping displacements of the optimized blades change little, the torsion angle of the blade tip is significantly reduced.

[0091] Table 2 Comparison of blade displacement and torsion angle

[0092]

[0093] The design method proposed in the present invention takes into account both the aerodynamic and structural performance of the airfoil. On the basis of improving the maximum lift-to-drag ratio of the airfoil, the airfoil moment coefficient, the airfoil section moment of inertia and the torsional moment of inertia are added to the optimization target. The moment of inertia is an important parameter for measuring the ability of the section to resist deformation. This optimization process not only improves the aerodynamic performance of the airfoil, but also enhances the structural stability of the blade. In addition, the present invention comprehensively considers the aerodynamic performance and structural strength of the airfoil of the blade cross section, and allocates different aerodynamic and structural targets to airfoils of different thicknesses from the root to the tip of the blade, so as to obtain an airfoil profile with better comprehensive performance, thereby enhancing the overall performance of the blade.

[0094] The advantages of the surrogate model-based airfoil optimization design process of the present invention are:

[0095] (1) A method for optimizing the airfoil design that comprehensively considers aerodynamic performance and structural characteristics is proposed. The maximum lift-to-drag ratio at the airfoil operating angle of attack is taken as the aerodynamic target, while the quadratic moment of inertia and torsional moment of inertia of the airfoil section are taken as structural targets. A genetic algorithm is used to find the optimal airfoil that takes both aerodynamic and structural performance into consideration. The maximum lift-to-drag ratio of the airfoil is improved and the moment coefficient fluctuation is reduced. When applied to the blade, the blade swing and torsion are reduced.

[0096] (2) When constructing the initial sample point set, spatial sampling is performed using the Latin hypercube sampling method optimized by the maximum and minimum distance criterion, which ensures that the sample points in the design space are evenly distributed, can better obtain the main features of the proxy model, and reduces the number of initial sample points.

[0097] (3) After constructing the initial Kriging proxy model, new samples are continuously added to the maximum value of the expected improvement of the model, model training is added, and continuous reconstruction is performed until a proxy model that meets the accuracy requirements is obtained. The point-adding criterion is used to greatly improve the optimization efficiency.

[0098] The contents described in the embodiments of the present invention are merely an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms described in the embodiments. The present invention can also make more deformations and modifications, which are all within the scope of protection required by the present invention.

Claims

1. A method for designing a wind turbine blade airfoil, characterized in that: The following steps are involved: Acquiring a sample point set of a blade airfoil, and calculating aerodynamic parameters and structural parameters of the blade airfoil; Establishing a proxy model of the blade airfoil according to the geometric parameters of the sample point set, the aerodynamic parameters of the blade airfoil and the structural parameters; The proxy model is optimized by using an optimization algorithm that takes the structural parameters into consideration to obtain an optimized airfoil.

2. The wind turbine blade airfoil design method according to claim 1, characterized in that: The step of obtaining a sample point set of a blade airfoil comprises the following steps: Performing geometric parameterization processing on the blade airfoil to obtain an optimization space; Sample points are extracted in the optimization space to form the sample point set.

3. The wind turbine blade airfoil design method according to claim 1, characterized in that: The calculating of the aerodynamic parameters and structural parameters of the blade airfoil comprises the following steps: Establishing a grid model of the blade airfoil, and performing mechanical analysis on the grid model to obtain the aerodynamic parameters; The geometric data of the cross section of the blade airfoil is obtained, and the structural parameters are calculated.

4. The wind turbine blade airfoil design method according to claim 1, characterized in that: The aerodynamic parameters include lift coefficient, drag coefficient and lift-to-drag ratio; and / or, The structural parameters include section moment of inertia and torsional moment of inertia.

5. The wind turbine blade airfoil design method according to claim 1, characterized in that: According to the geometric parameters of the sample point set, the aerodynamic parameters of the blade airfoil and the structural parameters, a proxy model of the blade airfoil is established, comprising the following steps: According to the geometric parameters, obtaining an input vector corresponding to each sample point in the sample point set; Acquiring a response value of each of the sample points according to the aerodynamic parameters and the structural parameters; A Kriging model of the blade airfoil is established according to the input vector and the response value.

6. The wind turbine blade airfoil design method according to claim 5, characterized in that: The wind turbine blade airfoil design method further comprises the following steps: Calculating an EI function based on the Kriging model obtained from the initial sample point set; The maximum value of the EI function is calculated by using the simplex method, and the sample point corresponding to the maximum value is added to the sample point set; Repeat the above steps and continue to add new sample points until the generated Kriging model meets the preset accuracy.

7. The wind turbine blade airfoil design method according to claim 1, characterized in that: The step of optimizing the proxy model by using an optimization algorithm that takes into account the structural parameters to obtain an optimized airfoil includes the following steps: An optimization objective function is established, and the maximum objective function is taken as the optimization target to obtain the optimized airfoil, wherein the objective function is: In the formula, C l With C d are the lift coefficient and drag coefficient of the blade airfoil, α1 and α2 are the two selected angles of attack, I x with I y is the section inertia moment of the blade airfoil cross section around the x-axis and the y-axis, J is the torsional inertia moment of the blade airfoil cross section around the centroid, w1, w2, w3 and w4 are weight coefficients, and w1+w2+w3+w4=1.

8. The wind turbine blade airfoil design method according to claim 7, characterized in that: The said establishing the optimization objective function, taking the maximum objective function as the optimization target, and obtaining the said optimized airfoil, further comprises the following steps: An optimization mathematical model is established, the optimization objective function is maximized as the optimization target, and the blade airfoil is optimized by using a global optimization algorithm. The optimization mathematical model is: Where: t max With t cmax are the positions of the maximum relative thickness and maximum thickness of the airfoil, respectively. Low and up represent the upper and lower limits of the value range. The superscript 0 represents the parameters of the reference airfoil. Setting constraints that the optimization objective needs to satisfy, wherein the constraints are an increase in lift coefficient, a decrease in drag coefficient, and an increase in sectional moment of inertia and torsional moment of inertia; Determine whether the optimization target satisfies the constraint condition. If so, output the optimized airfoil. If not, continue to add sample points until the constraint condition is satisfied.

9. The wind turbine blade airfoil design method according to claim 8, characterized in that: The wind turbine blade airfoil design method further comprises the following steps: If the error between the values ​​output by the mathematical optimization model calculated twice in succession is less than the convergence error, the optimization result is output and the optimization process ends.

10. The wind turbine blade airfoil design method according to claim 1, characterized in that: The wind turbine blade airfoil design method further comprises the following steps: Using CST parameterization method to parameterize the airfoil blade; and / or, Adopting an optimized Latin hypercube sampling method to obtain the sample point set; and / or, A grid topology model of the blade airfoil is established in ANSYS ICEM software to generate a structural grid, and boundary conditions are set in FLUENT software to perform flow field calculation and analysis.

Images