A method and device for predicting aerodynamic characteristics of wind turbine blades under dynamic stall

Abstract

The invention discloses a method and device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall. A high-fidelity model of aerodynamic characteristics of a wind turbine blade under dynamic stall is quickly constructed by using a small amount of dynamic stall wind tunnel test data. A static flow separation response model is constructed according to Kirchhoff flow theory and Newton interpolation method to calculate the dynamic stall characteristic transformation caused by dynamic flow separation in the B-L model; a B-L model optimization function is constructed under a given blade airfoil, and the optimal empirical time coefficient corresponding to the current wind turbine blade airfoil is calculated according to a small amount of dynamic stall wind tunnel test data; and a proxy model of the optimized B-L model is constructed by using a Kriging model. The uncertainty influence of different airfoils on the dynamic stall characteristics is fully considered, and the static flow separation calculation of the B-L model is improved to effectively improve the prediction accuracy of the dynamic stall characteristics. The high-fidelity proxy model can quickly obtain the dynamic stall aerodynamic characteristic response value of the wind turbine blade.

Description

Technical Field

[0001] The invention belongs to the technical field of aerodynamic characteristic analysis of wind turbine blades, and in particular relates to a method and a device for predicting aerodynamic characteristics of wind turbine blades under dynamic stall. Background Art

[0002] In order to maximize the capture of wind energy and reduce the load under high wind speed, existing large-scale horizontal axis wind turbines usually adopt variable speed and variable pitch control. During the actual operation of wind turbines, the adjustment of wind turbulence, yaw, rotor speed and blade pitch will cause the angle of attack of wind turbine blades to change rapidly, resulting in dynamic stall. Dynamic stall will change the original aerodynamic characteristics of the airfoil, thereby affecting the life and power generation of the wind turbine. Therefore, it is particularly important to accurately describe the dynamic stall characteristics. The BL (Beddoes-Leishman) model is a general semi-empirical engineering model for describing dynamic stall characteristics. In the traditional calculation process of the BL model, a piecewise exponential function is used to describe the static separation position relationship. For some airfoils, a large fitting error will occur, resulting in an inaccurate model. The empirical time coefficient used in the BL dynamic stall model is the fitting result for a certain airfoil. After testing, there is a large gap between the fitting empirical coefficients of different airfoils. If the empirical time coefficient is not re-calibrated, the model calculation will also be inaccurate. The BL dynamic stall model calculates the dynamic stall characteristics by solving a large number of differential equations. Although it does not take long for a single calculation, the purpose of this model is to simulate the aerodynamic load of wind turbines. During the simulation process, this model will be called a large number of times, which will affect the total simulation time. Summary of the invention

[0003] In order to solve the shortcomings of the prior art, the uncertain influence of different airfoils on the BL model is considered, the empirical time coefficient related to the airfoil characteristics is checked, the description form of the static flow separation function is modified, and the calculation accuracy of the BL model is effectively improved. At the same time, the purpose of rapid response to dynamic stall characteristics is achieved through the Kriging model. The present invention adopts the following technical solutions:

[0004] A method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall, comprising the following steps:

[0005] Step S1: based on the blade airfoil, obtaining corresponding static lift-drag aerodynamic characteristics and dynamic stall aerodynamic characteristics;

[0006] Step S2: constructing a flow separation response model based on static lift-drag aerodynamic characteristics, including the following steps:

[0007] Step S2.1: Calculate the normal force coefficient according to the static lift-drag aerodynamic characteristics and the corresponding angle of attack, and construct a linear function relationship based on the normal force coefficient and the angle of attack to obtain the function slope;

[0008] Step S2.2: Using Kirchhoff's flow theory, a functional relationship is established between the original static flow separation point and the normal force coefficient and its function slope to obtain the flow separation point at a given attack angle;

[0009] Step S2.3: within the effective attack angle range, the flow separation point is gradually reduced as the attack angle increases, and invalid results that do not conform to the law are removed;

[0010] Step S2.4: Determine the validity of the flow separation point. If not, jump to step S2.1 and adjust the linear function relationship. If valid, jump to step S2.5.

[0011] Step S2.5: Calculate using the Newton interpolation algorithm to obtain a static flow separation response model required in the BL dynamic stall model to calculate the dynamic flow separation process;

[0012] Step S3: Integrate the static flow separation response model into the BL dynamic stall model, replace the original flow separation function fitted by the piecewise exponential function, and obtain the dynamic stall aerodynamic characteristic response model;

[0013] Step S4: establishing optimization constraints of the empirical time coefficient based on the dynamic stall aerodynamic characteristics;

[0014] Step S5: constructing an objective function for minimizing the error between the dynamic stall aerodynamic characteristic response model and the empirical time coefficient (dynamic stall wind tunnel test data of the airfoil) under corresponding constraint conditions;

[0015] Step S6: solving the minimum optimization problem of the objective function to obtain the optimal empirical time coefficient of the corresponding blade airfoil;

[0016] Step S7: constructing a dynamic stall aerodynamic characteristics full-response high-fidelity model of the BL dynamic stall model after optimizing the empirical time coefficient.

[0017] Furthermore, in step S2.1, the static lift-drag aerodynamic characteristics include the lift coefficient C L and the drag coefficient C D , with the normal force coefficient C N , tangential force coefficient C T For different reference coordinate systems, packet transformation is performed to obtain two pairs of components representing aerodynamic characteristics:

[0018] C N =C L cos(a)+C D sin(a)

[0019] C T =C Lsin(a)-C D cos(a)

[0020] Where a represents the angle of attack.

[0021] Furthermore, in step S2.1, the normal force coefficient C is calculated by linear interpolation method. N = 0, the least square method is used to fit the normal force coefficient C within a certain angle range of a0 (±5°). N The linear function relationship of the function slope C is obtained Na .

[0022] Furthermore, in step S2.4, if it is invalid, jump to step S2.1, reduce the original fitting angle range (positive and negative 0.1° each), and recalculate the function slope C Na .

[0023] Furthermore, in step S2.2, the functional relationship is as follows:

[0024]

[0025] Among them, f N (a) represents the discrete function of the flow separation point at the angle of attack a, C N (a) represents the normal force coefficient corresponding to the attack angle a, C Na (a-a0) represents the angle of attack a and C N = The slope of the linear function formed by the angle of attack a0 when .

[0026] Furthermore, in step S2.4, if there is a flow separation point f>1, the current calculation is invalid, and the process jumps to step S2.1 and adjusts the linear function relationship; if there is no f>1, the current calculation is valid, and the process jumps to step S2.5.

[0027] Furthermore, in step S4, the constraints include:

[0028] Empirical time coefficient T f , its value range is T f >0,T f Characterize the hysteresis effect of unsteady dynamic stall flow separation relative to static separation;

[0029] Empirical time coefficient T p , its value range is T p >0,T p Characterize the hysteresis effect of the unsteady dynamic stall pressure response relative to the static pressure response on the wake separation effect and the airflow acceleration effect;

[0030] Empirical time coefficient T v, its value range is T v >0,T v It reflects that the accumulated vortex lift decreases exponentially with time.

[0031] Furthermore, in step S5, the objective function is:

[0032]

[0033] f(T)=f(T,w,k,c,a 0 ,a m ,a)

[0034] Where, f(T) represents the dynamic stall characteristics output by the BL dynamic stall model under a given input, Y r represents the dynamic stall characteristics during the increase of the angle of attack, Y d represents the dynamic stall characteristics during the process of angle of attack decreasing, n represents the number of samples during the process of angle of attack increasing, m represents the number of samples during the process of angle of attack decreasing, T=[T f ,T v ,T p ] represents the empirical time coefficient of the dynamic stall model and the dynamic stall aerodynamic characteristics related to the airfoil, w represents the angle of attack rotation frequency, k represents the reduction frequency, c represents the airfoil chord length, a 0 represents the average angle of attack, a m represents the pitch amplitude of the angle of attack, and a represents the angle of attack in the current state.

[0035] Furthermore, in step S7, each variable is equally spaced within a specified range to construct a high-dimensional input matrix, the angle of attack, rotation frequency, reduction frequency, average angle of attack, and angle of attack pitch amplitude range are set, and the high-dimensional input matrix is ​​input into the optimized BL dynamic stall model to obtain the corresponding dynamic stall characteristic response value, and a Kriging model is constructed based on the dynamic stall characteristic response value to obtain a dynamic stall characteristic response proxy model.

[0036] A device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall comprises a memory and one or more processors. The memory stores executable codes. When the one or more processors execute the executable codes, they are used to implement a method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall.

[0037] The advantages and beneficial effects of the present invention are:

[0038] The invention discloses a method and device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall. The static flow separation function calculation method thereof can more accurately express the relationship between the static flow separation point and the normal force coefficient. For different airfoils, a particle swarm algorithm is used to optimize the time empirical coefficient in the original BL model to reduce the model calculation error caused by the airfoil characteristics. By constructing a high-fidelity Kriging model of the BL model, the response speed of obtaining dynamic stall characteristics is greatly improved while ensuring that a small error is introduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flow chart of a method in an embodiment of the present invention.

[0040] Figure 2 1 is a graph showing the relationship between the static lift-drag coefficient and the angle of attack of the S814 airfoil in an embodiment of the present invention.

[0041] Figure 3 1 is a graph showing the relationship between the dynamic lift coefficient and the angle of attack of the S814 airfoil in an embodiment of the present invention.

[0042] Figure 4 It is a schematic diagram of the calculation comparison of the S814 airfoil flow separation response model before and after the improvement in the embodiment of the present invention.

[0043] Figure 5 It is a comparison diagram of the output results of the improved BL model and the original BL model in the embodiment of the present invention.

[0044] Figure 6a It is one of the partial result graphs output by the full response proxy model in the embodiment of the present invention.

[0045] Figure 6b This is the second partial result diagram of the full response proxy model output in the embodiment of the present invention.

[0046] Figure 6c This is the third partial result diagram of the full response proxy model output in the embodiment of the present invention.

[0047] Figure 6d This is the fourth partial result diagram of the full response proxy model output in the embodiment of the present invention.

[0048] Figure 6e This is the fifth partial result diagram of the full response proxy model output in the embodiment of the present invention.

[0049] Figure 6f This is the sixth partial result diagram of the full response proxy model output in the embodiment of the present invention.

[0050] Figure 7 Schematic diagram of the structure of the device in the embodiment of the present invention. DETAILED DESCRIPTION

[0051] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0052] Aiming at the rapid prediction of aerodynamic characteristics of large wind turbines, the present invention provides a method for predicting aerodynamic characteristics of wind turbine blades under dynamic stall, which considers the influence of different blade airfoils on dynamic stall characteristics and uses a high-fidelity proxy model to replace the original BL model to improve the speed of obtaining blade dynamic stall characteristics. Figure 1 As shown, the specific steps include:

[0053] Step S1: Obtain static lift-drag aerodynamic characteristics of a given blade airfoil, and partial dynamic stall aerodynamic characteristics data of a given blade airfoil measured by a wind tunnel experiment.

[0054] Specifically, the static lift-drag aerodynamic characteristics of a given blade airfoil include the corresponding static lift coefficient C at a specified angle of attack: L , drag coefficient C D .

[0055] Step S2: According to the static lift-drag aerodynamic characteristics of the blade airfoil, a flow separation response model is constructed using Kirchhoff flow theory and linear interpolation method, which specifically includes the following steps:

[0056] Step S2.1: Lift coefficient C L , drag coefficient C D and the normal force coefficient C N , tangential force coefficient C T For different references

[0057] The two pairs of components representing aerodynamic characteristics under the standard system have the corresponding coordinate transformation relationship as follows:

[0058] C N =C L cos(a)+C D sin(a)

[0059] C T =C L sin(a)-C D cos(a)

[0060] Calculate the normal force coefficient C N , using the normal force coefficient C N , the normal force coefficient C is calculated by linear interpolation N = 0, the angle of attack a0, the least squares method is used to fit the normal force coefficient C within the range of a0 ± 5° N The linear function relationship of the function slope C is obtained Na ;

[0061] Step S2.2: Using Kirchhoff's flow theory, transform the equations of the original static flow separation point and the normal force coefficient into

[0062] The functional relationship is:

[0063]

[0064] Substitute into the C calculated in step S2.1 N (a), C Na , C N , the discrete function f of the flow separation point at a given angle of attack is obtained N (a);

[0065] Step S2.3: within the effective angle of attack range, the flow separation point f gradually decreases as the angle of attack increases, and invalid results that do not conform to this rule are removed;

[0066] Step S2.4: Determine all flow separation points f obtained by calculation. If there is f>1, the current calculation is invalid, jump to step S2.1, reduce the original fitting range by 0.1° on both sides and recalculate C Na ; If f>1 does not exist, this calculation is valid and jump to the next step;

[0067] Step S2.5: Using the Newton interpolation algorithm, the static flow separation response model required in the BL calculation of dynamic flow separation is calculated.

[0068] Step S3: Integrate the above flow separation response model into the BL dynamic stall model, replace the original flow separation function fitted by the piecewise exponential function, and obtain the dynamic stall aerodynamic characteristic response model.

[0069] Step S4: Establish the constraints of the optimization problem:

[0070] (1) Empirical time coefficient T f Value range T f >0, where T f It characterizes the hysteresis effect of unsteady dynamic stall flow separation relative to static separation;

[0071] (2) Empirical time coefficient T p Value range T p >0, where T p It characterizes the hysteresis effect of the unsteady dynamic stall pressure response relative to the static pressure response on the wake separation effect and the airflow acceleration effect;

[0072] (3) Empirical time coefficient T v Value range T v >0, where T vIt reflects that the accumulated vortex lift decreases exponentially with time.

[0073] Step S5: Establishing an objective function C(T) for minimizing the error between the dynamic stall aerodynamic characteristic response model and the corresponding airfoil dynamic stall wind tunnel test data:

[0074] f(T)=f(T,w,k,c,a 0 ,a m ,a)

[0075]

[0076] Among them, the optimization target T = [T f ,T v ,T p ] represents the three time empirical coefficients of the dynamic stall model related to the airfoil, w is the angle of attack rotation frequency, k is the reduction frequency, c is the airfoil chord length, a 0 is the average angle of attack, a m is the pitch amplitude of the angle of attack, a is the angle of attack in the current state, f(T) is the dynamic stall characteristic of the BL model output under a given input, Y r =[y r1 ,y r2 ,...,y r3 ] represents the dynamic stall characteristics during the increase of the angle of attack measured in the wind tunnel experiment, Y d =[y d1 ,y d2 ,...,y d3 ] represents the dynamic stall characteristics during the process of reducing the angle of attack measured in the wind tunnel experiment.

[0077] The corresponding airfoil dynamic stall wind tunnel test data refers to the partial dynamic stall aerodynamic characteristics of the blade airfoil, including the rotation frequency w, reduction frequency k, average angle of attack a at a specified angle of attack a. 0 , angle of attack pitch amplitude a m The corresponding dynamic lift coefficient C L , drag coefficient C D .

[0078] Step S6: Using the optimization toolbox in MATLAB, the target minimum optimization problem is solved to obtain the optimal empirical time coefficient solution of the corresponding blade airfoil.

[0079] Specifically, the particle swarm optimization algorithm is used to solve the multi-objective optimization problem.

[0080] Step S7: using the Kriging model to construct a dynamic stall aerodynamic characteristics full-response high-fidelity model of the BL dynamic stall model after optimizing the empirical time coefficient;

[0081] Each variable takes points at equal intervals within a specified range to construct a high-dimensional input matrix. Set the angle of attack, rotation frequency, reduced frequency, mean angle of attack, and the range of angle of attack pitch amplitude. Input the high-dimensional input matrix into the optimized B-L model to obtain the dynamic stall characteristic response values corresponding to the inputs. Using the data obtained above, construct a Kriging model according to the DACE toolbox in MATLAB to obtain a surrogate model for the dynamic stall characteristic response.

[0082] In the embodiment of the present invention, each variable takes 100 points at equal intervals within a specified range to construct a high-dimensional input matrix, where the angle of attack range is -5° < a < 30°, the rotation frequency range is 0 Hz < w < 4 Hz, the reduced frequency range is 0 Hz < k < 0.2 Hz, the mean angle of attack range is 0° < a 0 < 20°, and the range of angle of attack pitch amplitude is 0° < a m < 20°. Input the high-dimensional input matrix into the optimized B-L model to obtain the dynamic stall characteristic response values corresponding to the inputs. Using the data obtained above, construct a Kriging model according to the DACE toolbox in MATLAB to obtain a surrogate model for the dynamic stall characteristic response.

[0083] To verify the effectiveness of the method of the present invention, the static and dynamic aerodynamic characteristic data of the S814 airfoil given by the National Renewable Energy Laboratory of the United States are adopted in the embodiment of the present invention, such as Figure 2 、 Figure 3 shown. Substitute it into the method of the present invention and compare it with the construction method of the original B-L model.

[0084] In the embodiment, according to the static lift-drag aerodynamic characteristics under the airfoil of the blade, a flow separation response model is constructed by using Kirchhoff flow theory and linear interpolation method. Compare it with the calculation method of the original B-L model as Figure 4 shown. The experimental empirical coefficients of the B-L model adopt the default values given in Table 1, and at the same time, they are used as the default initial values for the subsequent optimization of the time empirical coefficients.

[0085] Table 1 Empirical time coefficients of the B-L model

[0086]

[0087] The particle swarm optimization algorithm is used to optimize the empirical time coefficients of the B-L model. The results obtained by the method of the present invention are compared with the calculation results of the original B-L model as shown in Table 2. When fitting the experimental data of the dynamic aerodynamic characteristics of the S814 airfoil, the root mean square error of the improved B-L model fitting the experimental data is smaller. Compared with the original B-L model, this method can greatly improve the prediction accuracy. Some comparison results are as Figure 5The Kriging model is used to build a high-fidelity proxy model of the improved BL model. The model outputs some response results as shown in Figures 6a to 6f shown.

[0088] Table 2 Comparison results between the BL model and the improved BL model for S814 airfoil

[0089]

[0090] Corresponding to the aforementioned embodiment of a method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall, the present invention further provides an embodiment of a device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall.

[0091] See also Figure 7 An embodiment of the present invention provides a device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall, comprising a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, they are used to implement a method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall in the above embodiment.

[0092] An embodiment of a device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to the present invention can be applied to any device with data processing capability, and the device with data processing capability can be a device or apparatus such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, the processor of any device with data processing capability reads the corresponding computer program instructions in the non-volatile memory into the internal memory and runs them. From a hardware perspective, if Figure 7 As shown, it is a hardware structure diagram of any device with data processing capability where the device for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall of the present invention is located, except Figure 7 In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in which the apparatus in the embodiments is located may also include other hardware, generally based on the actual functions of the device with data processing capabilities, which will not be described in detail.

[0093] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0094] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

[0095] An embodiment of the present invention further provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, a method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall in the above embodiment is implemented.

[0096] The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device of any device with data processing capability, such as a plug-in hard disk, a smart media card (SMC), an SD card, a flash card, etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capability. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store data that has been output or is to be output.

[0097] The implementation method of this part is similar to that of the above method embodiment and will not be repeated here.

[0098] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the aerodynamic characteristics of a wind turbine blade under dynamic stall, characterized in that The steps include: Step S1: based on the blade airfoil, obtaining corresponding static lift-drag aerodynamic characteristics and dynamic stall aerodynamic characteristics; Step S2: constructing a flow separation response model based on static lift-drag aerodynamic characteristics, including the following steps: Step S2.1: Calculate the normal force coefficient according to the static lift-drag aerodynamic characteristics and the corresponding angle of attack, and construct a linear function relationship based on the normal force coefficient and the angle of attack to obtain the function slope; Step S2.2: Using Kirchhoff's flow theory, a functional relationship is established between the original static flow separation point and the normal force coefficient and its function slope to obtain the flow separation point at a given attack angle; Step S2.3: The flow separation points within the effective attack angle range gradually decrease with the increase of the attack angle, and the invalid flow separation points that do not conform to the gradual decrease with the increase of the attack angle are removed; Step S2.4: Determine the validity of the flow separation point. If not, jump to step S2.1 and adjust the linear function relationship. If valid, jump to step S2.

5. Step S2.5: Calculate using the interpolation algorithm to obtain a static flow separation response model required in the process of calculating dynamic flow separation using the Beddoes-Leishman dynamic stall model; Step S3: integrating the static flow separation response model into the Beddoes-Leishman dynamic stall model, replacing the flow separation function of the existing Beddoes-Leishman dynamic stall model fitted by the piecewise exponential function, and obtaining a dynamic stall aerodynamic characteristic response model; Step S4: Based on the dynamic stall aerodynamic characteristics, establish optimization constraints of the empirical time coefficient; the constraints include: Empirical time coefficient T f , its value range is T f >0,T f Characterize the hysteresis effect of unsteady dynamic stall flow separation relative to static separation; Empirical time coefficient T p , its value range is T p >0,T p Characterize the hysteresis effect of the unsteady dynamic stall pressure response relative to the static pressure response on the wake separation effect and the airflow acceleration effect; Empirical time coefficient T v , its value range is T v >0,T v It reflects that the accumulated vortex lift decreases exponentially with time; Step S5: constructing an objective function for minimizing the error between the dynamic stall aerodynamic characteristic response model and the empirical time coefficient under the corresponding constraint conditions; objective function: f(T)=f(T,w,k,c,a 0 ,a m ,a) Where, f(T) represents the dynamic stall characteristics of the Beddoes-Leishman dynamic stall model under a given input, Y r represents the dynamic stall characteristics during the increase of the angle of attack measured in the wind tunnel experiment, Y d represents the dynamic stall characteristics during the angle of attack reduction process measured by the wind tunnel experiment, n represents the number of samples during the angle of attack increase, m represents the number of samples during the angle of attack decrease, T = [T f ,T v ,T p ] represents the time experience coefficient, w represents the angle of attack rotation frequency, k represents the reduction frequency, c represents the airfoil chord length, a 0 represents the average angle of attack, a m represents the pitch amplitude of the angle of attack, and a represents the angle of attack in the current state; Step S6: Solve the minimum optimization problem of the objective function to obtain the optimal empirical time coefficient of the corresponding blade airfoil; Step S7: Construct a full-response high-fidelity model of the dynamic stall aerodynamic characteristics of the Beddoes-Leishman dynamic stall model after optimizing the empirical time coefficient.

2. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 1, characterized in that: In step S2.1, the static lift-drag aerodynamic characteristics include the lift coefficient C L and the drag coefficient C D , with the normal force coefficient C N , tangential force coefficient C T For different reference coordinate systems, coordinate transformation is performed to obtain two pairs of components representing aerodynamic characteristics: C N =C L cos(a)+C D sin (a) C T =C L sin(a)-C D cos (a) Where a represents the angle of attack.

3. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 1, characterized in that: In step S2.1, the normal force coefficient C is calculated by linear interpolation method. N = 0, the least square method is used to fit the attack angle a0 and the normal force coefficient C within a certain angle range of positive and negative a0. N The linear function relationship of the function slope C is obtained Na .

4. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 3, characterized in that: In step S2.4, if it is invalid, jump to step S2.1, reduce the original fitting angle range, and recalculate the function slope C Na .

5. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 1, characterized in that: In step S2.2, the functional relationship is as follows: Among them, f N (a) represents the discrete function of the flow separation point at the angle of attack a, C N (a) represents the normal force coefficient corresponding to the attack angle a, C Na (a-a0) represents the angle of attack a and C N = The slope of the linear function formed by the angle of attack a0 when .

6. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 5, characterized in that: In step S2.4, if there is a flow separation point f>1, the current calculation is invalid, jump to step S2.1, and adjust the linear function relationship; if there is no f>1, the current calculation is valid, jump to step S2.

5.

7. The method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to claim 1, characterized in that: In step S7, each variable is equally spaced within a specified range to construct a high-dimensional input matrix, the angle of attack, rotation frequency, reduction frequency, average angle of attack, and angle of attack pitch amplitude range are set, and the high-dimensional input matrix is ​​input into the optimized Beddoes-Leishman dynamic stall model to obtain the corresponding dynamic stall characteristic response value, and a Kriging model is constructed based on the dynamic stall characteristic response value to obtain a dynamic stall characteristic response proxy model.

8. A device for predicting aerodynamic characteristics of wind turbine blades under dynamic stall, characterized in that: It comprises a memory and one or more processors, wherein the memory stores executable codes, and when the one or more processors execute the executable codes, they are used to implement a method for predicting aerodynamic characteristics of a wind turbine blade under dynamic stall according to any one of claims 1 to 7.

Images