Optimal design method for fatigue life of wind turbine coating considering raindrop erosion

Abstract

The present invention discloses a method for optimizing the fatigue life of a wind turbine coating considering raindrop erosion, including: constructing a proxy model of the fatigue life of a wind turbine blade coating based on the Kriging method; constructing a probability distribution of wind field wind speed and rainfall intensity in a rain field considering the actual service environment; designing the speed of a wind turbine with the maximum power generation of the wind turbine as the objective function; constructing a constraint equation in combination with the fatigue life variation characteristics of the blade coating under different rainfall intensities; constructing an initial optimization point using the characteristics of the wind energy utilization coefficient for optimization; and obtaining the optimal speed of the wind turbine under different rainfall intensities and wind speeds by a traditional optimization calculation method. The method can effectively improve the fatigue life of the blade coating of the wind turbine while ensuring that the annual power generation of the wind turbine is reduced to a minimum, and provides a new idea for optimizing the fatigue life of the wind turbine coating.

Description

Technical Field

[0001] The invention relates to the field of wind turbine fatigue optimization, and in particular to a wind turbine coating fatigue life optimization design method taking raindrop erosion into consideration. Background Art

[0002] In recent years, in order to make better use of wind energy, the capacity of wind turbines has been increasing, resulting in the continuous increase in structural size parameters such as the diameter of wind turbine impellers. However, due to the long-term influence of factors such as rain corrosion, hail impact, and gravel friction on the leading edge of the blades, coupled with the high linear speed of the blades when the wind turbine is working, it is very easy to cause erosion of the leading edge of the blades, resulting in fatigue damage to the coating and reduced aerodynamic performance. Larger blade size means greater tip speed and contact area, which will lead to more serious erosion of the leading edge of the blades, affecting the power generation efficiency of the wind turbine. In addition, long-term erosion of the leading edge of the blades will lead to a decrease in production capacity, and even cause blade damage and accidents, affecting the normal operation and production economic benefits of the wind farm.

[0003] In order to reduce the fatigue damage caused by raindrop impact on the surface of wind turbine blades, existing engineering and research adopt methods such as modifying the surface coating material, increasing the thickness of the surface coating, regular maintenance and reducing the rotation speed, which can effectively prevent erosion phenomena such as raindrops and hail. This type of method is more effective in preventing erosion, but it requires multiple customized designs for different wind turbine blade surface structures and wind turbine service environments. In addition, for regular maintenance, frequent shutdown inspections may lead to waste of wind turbine production capacity in the absence of coating damage. As for the method of reducing the rotation speed, there is currently no method to quantitatively analyze it. Summary of the invention

[0004] The purpose of the present invention is to address the quantitative difficulties of the speed reduction method and propose an optimization design method for coating fatigue life that takes raindrop erosion into consideration. This method combines the construction of a proxy model for wind turbine blade coating fatigue life based on the Kriging method, considers the construction of the probability distribution of wind field wind speed and rain field rainfall intensity in the actual service environment, designs the wind turbine speed in combination with the maximum power generation of the wind turbine as the objective function, constructs constraint equations based on the fatigue life variation characteristics of blade coatings at different rainfall intensities, constructs the initial optimization point using the characteristics of wind energy utilization coefficients, and obtains the optimal speed of the wind turbine under different rainfall intensities and wind speeds through traditional optimization calculation methods, effectively improving the fatigue life of the wind turbine blade coating.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] S1: Establish a fatigue life proxy model for wind turbine blade coatings at different rotation speeds ω and different rainfall intensities I based on the Kriging method;

[0007] S1.1: Combined with the uniform sampling method, the fatigue life calculation method is used for different speeds ω and different rainfall intensities I to obtain the corresponding fan blade coating fatigue life t f ; Fatigue life calculation methods include but are not limited to: using finite element method and smooth particle fluid dynamics method to simulate raindrop impact on blades to obtain fatigue life, calculating fatigue life through water hammer-pressure equation and fatigue damage calculation formula, etc.;

[0008] S1.2: Reduce the magnitude difference of sample response values ​​by logarithmizing the fatigue life of fan blade coating;

[0009] S1.3: Construct the Kriging proxy model using the following formula;

[0010] t f (ω,I)=f T (ω,I)β+z(ω,I)

[0011]

[0012]

[0013] Among them, t f (ω,I) is the response function of the surrogate model, β is the regression coefficient, f(ω,I) is a second-order polynomial, and f T (ω,I)β is the approximate mean of the response, z(ω,I) is a Gaussian stationary random process with expectation E, variance Var, covariance Cov, σ z 2 is the variance, (ω i ,I i ) and (ω j ,I j ) are any two points in the training sample points, R(θ,ω i ,I i ,ω j ,I j ) is the correlation function equation with parameter θ;

[0014] S1.4: The fatigue life of the blade coating under a specific rotation speed and rainfall intensity is calculated by the method of S1.1 and the proxy model as the true value and the predicted value. The Kriging proxy model constructed in S1.3 is verified by combining the relative error. Encrypted sampling is performed around the points that do not meet the relative error requirements to obtain the corresponding fatigue life t f , and return to S1.2 to rebuild the proxy model;

[0015] S2: Construct the probability distribution of wind speed in the wind field and rainfall intensity in the rain field considering the actual service environment;

[0016] S2.1: Calculate the probability of a rainfall event P based on the total rainfall duration data for the whole year rain ;

[0017] S2.2: Based on the actual measured rainfall intensity frequency data, the maximum likelihood estimation method is used to fit the log-normal distribution f I The position parameters and size parameters of (I) are as follows:

[0018]

[0019] In the formula, f I (I) represents the probability of occurrence of rainfall intensity I in the log-normal distribution, μ I and σ I Location and scale parameters for the lognormal distribution.

[0020] S2.3: Combining the results of S2.1 and S2.2, the following formula is used to calculate the probability of occurrence of different rainfall intensities P I (I i )

[0021]

[0022] Where P I (I i ) is the rainfall intensity I i The probability distribution expression of occurrence, P rain is the probability of a rainfall event, and ΔI is the rainfall intensity interval.

[0023] S2.4: Based on the actual measured wind speed frequency data, the maximum likelihood estimation method is used to fit the Weibull distribution f v The scale parameter and shape parameter of (v) are as shown in the following formula:

[0024]

[0025]

[0026] Where C and k are the scale parameter and shape parameter of the Weibull distribution. v (v i ) is the wind speed v i Occurrence probability distribution expression.

[0027] S3: Establish the objective function f that maximizes the power generation of the wind turbine AEP (ω) Design the speed of wind turbine generator:

[0028]

[0029] In the formula, 8760 represents the number of hours in a year, N0 represents the designed service life of the fan (years), and P r (v j , ω ij ) is the wind speed v j and fan speedω ij Wind turbine power generation under I (I i ) is the rainfall intensity I i The probability of occurrence, P v (v j ) is the wind speed v j The probability of occurrence of each rainfall intensity is i With wind speed v j Fan speed ω ij As a component of the design variable, the expression of the design variable ω is:

[0030] ω=[ω 00 ,ω 01 ,...,ω 0(n-1) ,ω 0n ,ω 10 ,ω 11 ,...,ω m(n-1) ,ω mn ].

[0031] S3.1: According to Betz theory, the formula for calculating the power generated by a wind turbine from wind energy is:

[0032]

[0033] Where ρ is the density of air (kg / m3), which is generally 1.25kg / m 3 , R is the fan impeller radius, λ is the tip speed ratio, β is the pitch angle (determined by the fan variable pitch system according to the wind speed), C p (λ,β) is the calculation formula of wind energy utilization coefficient. Different C is selected according to different types of wind turbines. p Calculation formula.

[0034] The calculation formula of blade tip speed ratio λ is:

[0035]

[0036] Where ω is the fan speed (r / min)

[0037] S4: Constraint equations are constructed based on the variation characteristics of blade coating life under different rainfall intensities;

[0038] (1) The range of the design variable ω cut-in ≤ω≤ω rated, represents the cut-in speed of the wind turbine ω cut-in With rated speedω rated Set the upper and lower bounds for the fan speed design variable.

[0039] (2) In order to prevent the blade coating fatigue damage from reaching the design threshold, the constraint condition expression is set:

[0040]

[0041] In the formula, 52560 represents the number of 10-minute cycles in a year, and D 10 (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue damage caused by running for 10 minutes is expressed as:

[0042]

[0043] Where, t f (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue life of the wind turbine blade coating during operation (h) can be quickly calculated using the wind turbine blade coating fatigue life proxy model built by S1. f (I i ,ω ij ).

[0044] (3) The fatigue life of blade coating decreases rapidly exponentially with the increase of rainfall intensity. Therefore, when the wind speed is the same, as the rainfall intensity increases, the fan speed should be reduced to ensure that the fatigue life of blade coating is not too short. Based on this characteristic, the constraint expression is set to narrow the search space of high-dimensional problems:

[0045]

[0046] In the formula, ω (i+1)j is the rainfall intensity I i+1 With wind speed v j The fan speed under ij is the rainfall intensity I i With wind speed v j The fan speed at the bottom.

[0047] S5.1: Since the overall design space is large, in order to avoid falling into the local optimum and reduce the number of optimization search steps, it is necessary to select an optimization initial point that is close to the optimal point in the design space. According to the characteristics of wind energy utilization coefficient, under the same wind speed and pitch angle, the wind turbine power increases continuously with the increase of wind turbine speed. When the wind turbine power reaches the maximum value, increasing the wind turbine speed will cause the wind turbine power to decrease. For different wind speeds v j , there is always an optimal speed ω that maximizes the fan power 0_best_j , use the optimization method to calculate the speed within the cut-in speed and rated speed range, each wind speed v j The optimal speed ω that maximizes the fan power 0_best_j , the optimization model is as follows:

[0048] Find ω 0_best_j

[0049] Min C p (ω 0_best_j )

[0050] ST ω 切入 ≤ω 0_best_j ≤ω 额定

[0051] S5.2: Different rainfall intensities I i With wind speed v j The design variable fan speed ω ij The optimization initial value is set to ω 0_best_j , the overall expression is as follows:

[0052]

[0053] Among them, different wind speed v j The corresponding maximum power fan speed ω 0_best_j To optimize the initial point;

[0054] S6: According to the method described in S4, the wind turbine speed optimization model considering raindrop erosion is obtained as follows

[0055] Find ω=[ω 00 ,ω 01 ,...,ω 0(n-1) ,ω 0n ,ω 10 ,ω 11 ,...,ω m(n-1) ,ω mn ]

[0056] Max

[0057] ST

[0058]

[0059] ω 切入 ≤ω≤ω 额定

[0060] By using the optimization initial point calculated in S5 and solving the problem using the solver in MATLAB, the optimal solution corresponding to the design variables can be obtained.

[0061] Furthermore, in S1.1, the stress distribution of the wind turbine blade coating is obtained by using the finite element analysis method through ABAQUS software.

[0062] Furthermore, in S6, a sequential quadratic programming solver is used for solving the problem.

[0063] The beneficial effects of the present invention are as follows:

[0064] (1) The blade coating fatigue life calculation framework and the Kriging method are combined to construct a blade coating fatigue life proxy model to achieve rapid calculation of the blade coating fatigue life under different rainfall intensities and wind turbine speeds.

[0065] (2) Construct a rainfall intensity probability model and a wind speed probability model to characterize the duration information of wind speed and rainfall intensity in the service environment.

[0066] (3) A speed optimization method considering raindrop erosion is proposed. Constraints are constructed based on the fatigue life variation characteristics of the blade coating to narrow the search space. The initial speed optimization point is constructed based on the wind energy utilization coefficient characteristics of the wind turbine. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 is a flow chart of the method of the present invention;

[0068] Figure 2 This is a schematic diagram of the calculation framework for blade coating fatigue life;

[0069] Figure 3a and Figure 3b They are the comparison results of the proxy model predictions of different rainfall intensities at a rotation speed of 9.5493r / min before and after encrypted sampling;

[0070] Figure 4 It is a schematic diagram of the probability mass function of each rainfall intensity during the wet period;

[0071] Figure 5 It is a schematic diagram of the probability mass function of each wind speed;

[0072] Figure 6 It is a graph of wind energy utilization coefficient under different pitch angles;

[0073] Figure 7 It is the fatigue life variation diagram during the iteration process;

[0074] Figure 8 This is a comparison chart of fan speed before and after optimization. DETAILED DESCRIPTION

[0075] The present invention will be described in detail below based on the accompanying drawings and preferred embodiments, and the purpose and effects of the present invention will become more clear. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0076] The wind turbine blade coating fatigue analysis method considering raindrop erosion of the present invention uses the Kriging method to realize the construction of a proxy model for the fatigue life of the wind turbine blade coating; constructs the probability distribution of the wind field and the rain field considering the actual service environment; designs the wind turbine speed with the maximum power generation of the wind turbine as the objective function; constructs the constraint equation according to the fatigue life variation characteristics of the blade coating under different rainfall intensities; constructs the initial optimization point using the wind energy utilization coefficient characteristics for optimization, and performs optimization analysis on the 5MW wind turbine of the National Renewable Energy Laboratory (NREL) of the United States. The specific flow chart is as follows Figure 1 As shown, the specific steps include:

[0077] S1: Establish a fatigue life proxy model for wind turbine blade coatings at different rotation speeds ω and different rainfall intensities I based on the Kriging method;

[0078] S1.1: Combined with the uniform sampling method, the fatigue life calculation method is used for different speeds ω and different rainfall intensities I to obtain the corresponding fan blade coating fatigue life t f The process of this method is as follows Figure 2 As shown in the figure, a random rain field is constructed according to the rainfall intensity I. The stress field of raindrops hitting the blades is calculated in ABAQUS, and the fatigue life is calculated by combining the crack extension in the initiation and development stages of the coating fatigue. In order to control the amount of calculation, a finite element model is constructed for the blade tip panel, and the relevant properties are shown in Table 1 below.

[0079] Table 1 Material properties of each layer of the composite panel

[0080]

[0081] S1.2: Reduce the magnitude difference of sample response values ​​by logarithmizing the fatigue life of the fan;

[0082] S1.3: Construct the Kriging proxy model using the following formula;

[0083] t f (ω,I)=f T(ω,I)β+z(ω,I)

[0084]

[0085]

[0086] Among them, t f (ω,I) is the response function of the surrogate model, β is the regression coefficient, f(ω,I) is a second-order polynomial, and f T (ω,I)β is the approximate mean of the response, z(ω,I) is a Gaussian stationary random process with expectation E, variance Var, covariance Cov, σ z 2 is the variance, (ω i ,I i ) and (ω j ,I j ) are any two points in the training sample points, R(θ,ω i ,I i ,ω j ,I j ) is the correlation function equation with parameter θ;

[0087] S1.4: The fatigue life of the blade coating under a specific rotation speed and rainfall intensity is calculated by the method of S1.1 and the proxy model as the true value and the predicted value. The Kriging proxy model constructed in S1.3 is verified by combining the relative error. The surrounding areas of the points that do not meet the relative error requirements are encrypted and the corresponding fatigue life t is obtained. f , and return to S1.2 to rebuild the proxy model. The comparison results of the proxy model predictions of different rainfall intensities at a speed of 9.5493r / min before and after encrypted sampling are as follows: Figure 3a and 3b As shown in the figure, it can be seen that the predicted values ​​of blade coating fatigue life under 20 rainfall intensities are well matched with the true values, indicating that the blade coating fatigue life proxy model is more credible.

[0088] S2: Construct the probability distribution of wind speed in the wind field and rainfall intensity in the rain field considering the actual service environment;

[0089] S2.1: Calculate the probability of a rainfall event P based on the total rainfall duration data for the whole year rain ;

[0090] S2.2: Based on the actual rainfall intensity frequency data measured in the Miami area of ​​the United States, the maximum likelihood estimation method is used to fit the log-normal distribution f I The position parameters and size parameters of (I) are as follows:

[0091]

[0092] In the formula, f I (I) represents the probability of occurrence of rainfall intensity I in the log-normal distribution, μ I and σ I The location parameter and scale parameter of the lognormal distribution, and the probability mass function of each rainfall intensity during the wet period are fitted as follows: Figure 4 shown.

[0093] S2.3: Combining the results of S2.1 and S2.2, the following formula is used to calculate the probability of occurrence of different rainfall intensities P I (I i )

[0094]

[0095] Where P I (I i ) is the rainfall intensity I i The probability distribution expression of occurrence, P rain is the probability of a rainfall event, and ΔI is the rainfall intensity interval.

[0096] S2.4: Based on the wind speed frequency data actually measured in various regions of the United States, the maximum likelihood estimation method is used to fit the Weibull distribution f v The scale parameter and shape parameter of (v) are as shown in the following formula:

[0097]

[0098]

[0099] Where C and k are the scale parameter and shape parameter of the Weibull distribution. v (v i ) is the wind speed v i The probability distribution expression of occurrence is as follows: Figure 5 shown.

[0100] S3: Establish the objective function f that maximizes the power generation of the wind turbine AEP (ω) Design the speed of wind turbine generator:

[0101]

[0102] In the formula, 8760 represents the number of hours in a year, N0 represents the designed service life of the fan (years), and P r (v j , ω ij ) is the wind speed v j and fan speedω ij Wind turbine power generation under I (Ii ) is the rainfall intensity I i The probability of occurrence, P v (v j ) is the wind speed v j The probability of occurrence of each rainfall intensity is i With wind speed v j Fan speed ω ij As a component of the design variable, the expression of the design variable ω is:

[0103] ω=[ω 00 ,ω 01 ,...,ω 0(n-1) ,ω 0n ,ω 10 ,ω 11 ,...,ω m(n-1) ,ω mn ];

[0104] The wind turbine used is a 5MW wind turbine from the National Renewable Energy Laboratory (NREL) of the United States. Its parameters are shown in Table 2 below.

[0105] Table 2 NREL's 5MW wind turbine performance parameters

[0106]

[0107] S3.1: According to Betz theory, the formula for calculating the power generated by a wind turbine from wind energy is:

[0108]

[0109] Where ρ is the density of air (kg / m3), which is generally 1.25kg / m 3 , R is the fan impeller radius, λ is the tip speed ratio, β is the pitch angle (determined by the fan variable pitch system according to the wind speed), C p (λ,β) is the calculation formula for wind energy utilization coefficient.

[0110] The calculation formula of blade tip speed ratio λ is:

[0111]

[0112] Where ω is the fan speed (r / min)

[0113] S3.2: The wind energy utilization coefficient calculation formula C used in this case p (λ,β) is as follows

[0114]

[0115]

[0116] S4: Construct constraint equations based on the life span variation characteristics of blade coatings at different rainfall intensities;

[0117] (1) The range of the design variable ω cut-in ≤ω≤ω rated , represents the cut-in speed of the wind turbine ω cut-in , is 6.9 r min -1 With rated speedω rated , setting the upper and lower bounds of the fan speed design variable to 12.1 r min-1.

[0118] (2) In order to prevent the blade coating fatigue damage from reaching the design threshold, the constraint condition expression is set:

[0119]

[0120] In the formula, 52560 represents the number of 10-minute cycles in a year, and D 10 (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue damage caused by running for 10 minutes is expressed as:

[0121]

[0122] Where, t f (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue life of the wind turbine blade coating during operation (h) can be quickly calculated using the wind turbine blade coating fatigue life proxy model built by S1. f (I i ,ω ij ).

[0123] (3) The characteristic of blade coating fatigue life is that with the increase of rainfall intensity, the blade coating fatigue life decreases exponentially. Therefore, when the wind speed is the same, as the rainfall intensity increases, the fan speed should be reduced to ensure that the blade coating fatigue life is not too short. Based on this characteristic, the constraint expression is set to narrow the search space of the high-dimensional problem:

[0124]

[0125] In the formula, ω (i+1)j is the rainfall intensity I i+1 With wind speed v j The fan speed underij is the rainfall intensity I i With wind speed v j The fan speed at the bottom.

[0126] S5.1: Since the overall design space is large, in order to avoid falling into the local optimum and reduce the number of optimization search steps, it is necessary to select an optimization initial point that is close to the optimal point in the design space. According to the characteristics of wind energy utilization coefficient, under the same wind speed and pitch angle, the wind turbine power increases continuously with the increase of wind turbine speed. When the wind turbine power reaches the maximum value, increasing the wind turbine speed will cause the wind turbine power to decrease, such as Figure 6 As shown. For different wind speeds v j , there is always an optimal speed ω that maximizes the fan power 0_best_j , use the optimization method to calculate the speed within the cut-in speed and rated speed range, each wind speed v j The optimal speed ω that maximizes the fan power 0_best_j , the optimization model is as follows:

[0127] Find ω 0_best_j

[0128]

[0129] Min

[0130]

[0131] ST ω 切入 ≤ω 0_best_j ≤ω 额定

[0132] S5.2: Different rainfall intensities I i With wind speed v j The design variable fan speed ω ij The optimization initial value is set to ω 0_best_j , the overall expression is as follows:

[0133]

[0134] Among them, different wind speed v j The corresponding maximum power fan speed ω 0_best_j To optimize the initial point;

[0135] S6: According to the method described in S4, the wind turbine speed optimization model considering raindrop erosion is obtained as follows

[0136] Find ω=[ω 00 ,ω 01 ,...,ω 0(n-1) ,ω0n ,ω 10 ,ω 11 ,...,ω m(n-1) ,ω mn ]

[0137] Max

[0138] ST

[0139]

[0140] ω 切入 ≤ω≤ω 额定

[0141] By using the S5 calculation to get the optimization initial point, and using the solver in MATLAB to solve the problem, the optimal solution corresponding to the design variables can be obtained. Figure 7 As shown in the figure, it can be seen that as the optimization proceeds, the fatigue life of the blade coating continues to increase, and the fatigue life of the blade coating finally converges to the preset service life at 20 years, and the optimization process converges. Figure 8 As shown in the figure, it can be seen that when the rainfall intensity and wind speed are high, the fan speed drops significantly.

[0142] The annual energy production (AEP) and fatigue life of wind turbines before and after optimization are compared, as shown in Table 3. The wind turbine speed optimization method considering raindrop erosion increases the fatigue life of blade coating from 7.8393 years to 20 years, while the annual power generation AEP only decreases by 4094.92kW·h. This shows that the optimization method proposed in this paper can better balance the influence of wind speed on wind turbine power generation and the influence of rainfall intensity on the fatigue life of wind turbine blade coating through wind turbine speed, and maximize the annual energy production (AEP) of wind turbines while ensuring that the fatigue life of wind turbine blade coating reaches the designed service life.

[0143] Table 3 Annual power generation of wind turbines and coating fatigue life before and after optimization

[0144]

[0145] This example effectively improves the fatigue life of the wind turbine blade coating through the optimization method of the present invention, combined with the actual operation of the wind turbine, and ensures that the annual power generation is not significantly reduced. It can be understood by those of ordinary skill in the art that the above is only a preferred example of the invention and is not intended to limit the invention. Although the invention has been described in detail with reference to the aforementioned examples, it is still possible for those of ordinary skill in the art to modify the technical solutions recorded in the aforementioned examples, or to make equivalent substitutions for some of the technical features therein. All modifications, equivalent substitutions, etc. made within the spirit and principles of the invention should be included in the scope of protection of the invention.

Claims

1. A wind turbine coating fatigue life optimization design method considering raindrop erosion, characterized in that: The method comprises the following steps: Step S1: establishing a fatigue life proxy model of wind turbine blade coating under different rotation speeds ω and different rainfall intensities I based on Kriging method; Step S2: construct the probability distribution of wind speed in the wind field and rainfall intensity in the rain field considering the actual service environment; Step S3: Establish an objective function f that maximizes the power generation of the wind turbine AEP (ω) Design the speed of wind turbine generator: In the formula, 8760 represents the number of hours in a year, N0 represents the designed service life of the fan, and P r (v j ,ω ij ) is the wind speed v j and fan speedω ij The wind turbine power generation under I (I i ) is the rainfall intensity I i The probability of occurrence, P v (v j ) is the wind speed v j The probability of occurrence of each rainfall intensity is i With wind speed v j Fan speed ω ij As a component of the design variable, the expression of the design variable ω is: ω=[ω 00 ,oh 01 ,...,oh 0(n-1) ,oh 0n ,oh 10 ,oh 11 ,...,oh m(n-1) ,oh mn ]; According to Betz theory, the formula for calculating the power generated by a wind turbine from wind energy is: Where ρ is the density of air, which is 1.25 kg / m 3 , R is the fan impeller radius, λ is the tip speed ratio, β is the pitch angle, which is determined by the fan variable pitch system according to the wind speed, C p (λ,β) is the calculation formula of wind energy utilization coefficient. Different C is selected according to different types of wind turbines. p Calculation formula; The calculation formula of blade tip speed ratio λ is: Where, ω is the fan speed, the unit is revolutions per minute, r / min; Step S4: construct constraint equations based on the blade coating life variation characteristics under different rainfall intensities, specifically: The range of the design variable ω cut-in ≤ω≤ω rated , represents the cut-in speed of the wind turbine ω cut-in With rated speedω rated Set the upper and lower bounds of the fan speed design variable; To prevent blade coating fatigue damage from reaching the design threshold, constraints are set; Since the fatigue life of blade coating decreases exponentially with the increase of rainfall intensity, the fatigue life of blade coating can be guaranteed not to be too short by reducing the speed of wind turbine as the rainfall intensity increases at the same wind speed. Based on this characteristic, the constraint expression is set to reduce the search space of high-dimensional problem. Step S5: Since the overall design space is large, in order to avoid falling into the local optimum and reduce the number of optimization search steps, an optimization initial point close to the optimal point is selected in the design space. According to the characteristics of wind energy utilization coefficient, under the same wind speed and pitch angle, the wind turbine power increases with the increase of wind turbine speed; when the wind turbine power reaches the maximum value, increasing the wind turbine speed will cause the wind turbine power to decrease; for different wind speeds v j , there is always an optimal speed ω that maximizes the fan power 0_best_j , use the optimization method to calculate the speed within the cut-in speed and rated speed range, each wind speed v j The optimal speed ω that maximizes the fan power 0_best_j , the optimization model is as follows: Find ω 0_best_j Min C p (oh 0_best_j ), ST oh 切入 ≤ω 0_best_j ≤ω 额定 Different rainfall intensities I i With wind speed v j The design variable fan speed ω ij The optimization initial value is set to ω 0_best_j , the overall expression is as follows: Among them, different wind speed v j The corresponding maximum power fan speed ω 0_best_j To optimize the initial point; Step S6: According to the method described in step S4, a wind turbine speed optimization model considering raindrop erosion is obtained, as shown below: In the formula, 52560 represents the number of 10-minute cycles in a year, and D 10 (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue damage caused by running for 10 minutes is expressed as: Where, t f (I i ,ω ij ) indicates that when the rainfall intensity is I i The lower fan speed is ω ij The fatigue life limit of the wind turbine blade coating during operation can be quickly calculated by the wind turbine blade coating fatigue life proxy model built by S1. f (I i ,ω ij ); The optimization initial point is obtained by calculation in step S5, and the problem is solved by using the solver in MATLAB to obtain the optimal solution corresponding to the design variables.

2. The wind turbine coating fatigue life optimization design method considering raindrop erosion according to claim 1 is characterized in that: The step S1 is specifically as follows: S1.1: Combined with the uniform sampling method, the fatigue life calculation method is used for different speeds ω and different rainfall intensities I to obtain the corresponding fan blade coating fatigue life t f ; fatigue The life calculation methods include: using the finite element method and smooth particle fluid dynamics method to simulate raindrops hitting blades to obtain fatigue life, and calculating fatigue life through the water hammer-pressure equation and fatigue damage calculation formula; S1.2: Reduce the magnitude difference of sample response values ​​by logarithmizing the fatigue life of fan blade coating; S1.3: Construct the Kriging proxy model using the following formula; t f (ω,I)=f T (ω,I)β+z(ω,I), Among them, t f (ω,I) is the response function of the surrogate model, β is the regression coefficient, f(ω,I) is a second-order polynomial, and f T (ω,I)β is the approximate mean of the response, z(ω,I) is a Gaussian stationary random process with expectation E, variance Var, covariance Cov, σ z 2 is the variance, (ω i ,I i ) and (ω j ,I j ) are any two points in the training sample points, R(θ,ω i ,I i ,ω j ,I j ) is the correlation function equation with parameter θ; S1.4: The fatigue life of the blade coating under the rotation speed and rainfall intensity is calculated by the method of S1.1 and the proxy model as the true value and the predicted value respectively. The Kriging proxy model constructed in S1.3 is verified by combining the relative error. The sampling is encrypted around the points that do not meet the relative error requirements to obtain the corresponding fatigue life t f , and return to S1.2 to rebuild the proxy model.

3. The wind turbine coating fatigue life optimization design method considering raindrop erosion according to claim 2 is characterized by: In S1.1, the stress distribution of the wind turbine blade coating is obtained by using the finite element analysis method through ABAQUS software.

4. The method for optimizing the fatigue life of a wind turbine coating considering raindrop erosion according to claim 1 is characterized in that: The step S2 is specifically as follows: S2.1: Calculate the probability of a rainfall event P based on the total rainfall duration data for the whole year rain ; S2.2: Based on the actual measured rainfall intensity frequency data, the maximum likelihood estimation method is used to fit the log-normal distribution f I The position parameters and size parameters of (I) are as follows: In the formula, f I (I) represents the probability of occurrence of rainfall intensity I in the log-normal distribution, μ I and σ I Location and scale parameters of the lognormal distribution; S2.3: Combining the results of S2.1 and S2.2, the following formula is used to calculate the probability of occurrence of different rainfall intensities P I (I i ) Where P I (I i ) is the rainfall intensity I i The probability distribution expression of occurrence, P rain is the probability of occurrence of rainfall events, ΔI is the rainfall intensity interval; S2.4: Based on the actual measured wind speed frequency data, the maximum likelihood estimation method is used to fit the Weibull distribution f v The scale parameter and shape parameter of (v) are shown in the following formula: Where C and k are the scale parameter and shape parameter of the Weibull distribution, P v (v i ) is the wind speed v i Occurrence probability distribution expression.

5. The method for optimizing the fatigue life of a wind turbine coating considering raindrop erosion according to claim 1 is characterized in that: In the step S6, a sequential quadratic programming solver is used to solve the problem.

Images