Wind turbine tower reliability optimization method based on probability checking safety factor

Abstract

The application discloses a wind turbine tower reliability optimization method based on a probability checking safety coefficient, belongs to the field of tower reliability optimization design, and divides the safety coefficient checking into uncertain variable design values and characteristic values, constructs a multi-target reliability optimization design problem, constructs a mapping relationship between the uncertain variable design values and reliability, combines fixed quantile characteristic values, establishes a safety coefficient library available under different reliabilities, adopts a Wasserstein distance to evaluate the change degree of environmental load and guide safety coefficient selection in view of dynamic environmental load and different reliability design requirements, converts a complex reliability optimization problem into a deterministic optimization model for fast and efficient solution by matching the characteristic value quantile of the environmental parameters and the safety coefficient of the target reliability, obtains tower structure design variables, saves the calculation cost and design redundancy, and uses the design variables for the tower, which can improve the reliability of the tower under different service environments.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of tower reliability optimization design, and particularly relates to a wind turbine tower reliability optimization method based on a probability checking safety coefficient. BACKGROUND

[0002] There are various uncertain factors in the whole life cycle of wind turbine production, operation and service, which have a significant impact on the safety of the structure. In order to ensure the safety of the wind turbine, the safety coefficient is often used in engineering to resist the influence of various uncertainties. The safety coefficient method is widely used in wind turbine tower design due to its intuitive, easy to understand and convenient to use. However, as the complexity of the service environment of the wind turbine gradually increases, the design requirements for reliability are more diversified, and the safety coefficient system based on empirical criteria or simplified theoretical assumptions has been difficult to meet the design requirements. The overly conservative safety coefficient will cause serious design redundancy and waste of production cost. SUMMARY

[0003] In order to solve the problems of the prior art, realize the combination of multi-objective optimization design and safety coefficient checking under load uncertainty, and improve the reliability of the wind turbine tower under different service environments, the present application adopts the following technical solutions:

[0004] The wind turbine tower reliability optimization method based on the probability checking safety coefficient, the optimization process is as follows:

[0005] The structural parameters of the tower are taken as the determined design variables, the uncertain factors acting on the tower are taken as the uncertain variables, and the target reliability interval of the optimization target is determined according to the interval of the tower failure probability;

[0006] Based on the structural optimization target of the tower, the optimization target is to minimize the steel consumption of the tower, and in the case of using the same material steel, the design target is equivalent to minimizing the volume of the tower. In the target reliability interval, the maximum possible point of the reliability corresponding to the uncertain variable is obtained;

[0007] The design value of the uncertain variable corresponding to the maximum possible point under the reliability is selected; according to the distribution of the uncertain variable, a fixed quantile is selected as a characteristic value; through the design value and the characteristic value, the safety coefficient under different reliabilities is obtained;

[0008] The target reliability of different working conditions and load distribution is calculated ; combined with the reliability offset and the target reliability , the corresponding safety coefficient is selected to be integrated into the structural optimization of the tower to generate the optimal design variable corresponding to the adjusted arbitrary load distribution and different reliabilities.

[0009] Further, the tower buckling failure mode is obtained according to the stress and stress resistance of the tower, and the failure constraint is constructed by the tower buckling failure mode , the target reliability optimization is performed based on the failure constraint and the optimization target of the tower to obtain the maximum possible point of the uncertainty variable corresponding to different reliabilities in the target reliability interval.

[0010] Further, the fixed quantile is selected as the current environmental characteristic value according to the distribution of the uncertainty variable determined by the environment ; by measuring the distance between the load distribution when the safety factor is checked and the load distribution under the new environment, the weight coefficient is multiplied by the sensitivity at the target reliability , and the ratio of the product to the current environmental characteristic value is the reliability offset required by the change of the environmental load . In different environments, the distribution of the load may change, and the safety factor can eliminate the influence of this load uncertainty.

[0011] Further, the sensitivity is obtained by obtaining the most likely point corresponding to the target reliability in the standard normal space, and the square of the load partial derivative of the failure constraint at the most likely point is divided by the square of the norm of the gradient of the failure constraint at the most likely point.

[0012] Further, based on the characteristic value of the load variable under the new distribution, the material characteristic characteristic value, and the safety factor corresponding to the adjusted target reliability, the failure constraint is adjusted.

[0013] Further, the uncertainty variable includes material buckling strength , uncertainty factor and load, the uncertainty factor includes aeroelastic uncertainty factor , experimental uncertainty factor , aerodynamic uncertainty factor , structural uncertainty factor , the load includes the force and the bending moment received by the tower, and specifically includes the axial force , tangential bending moment , tangential force , axial bending moment , and the distribution parameter corresponding to the load is the load distribution parameter.

[0014] Further, the design variable is the tower thickness, the optimization objective is the minimum tower volume, the tower buckling failure mode is the difference between the tower stress resistance and stress, the stress resistance is obtained by multiplying the stress buckling reduction coefficient by the material buckling strength, the stress is obtained by multiplying the sum of the force received by the tower by the tower circumference and thickness, and the bending moment by the tower area and thickness, and then multiplying the sum by the uncertainty factor.

[0015] Further, the safety factor includes the safety factor of the material buckling strength, the safety factor of the force received by the tower, and the safety factor of the bending moment; the safety factor is integrated into the structural optimization of the tower, and the material buckling strength, the force received by the tower, and the bending moment in the tower buckling failure mode are multiplied by the respective safety factors.

[0016] Further, for the uncertainty variable related to the structural strength, the 5% quantile is selected as the characteristic value to ensure the design reliability; for the uncertainty variable related to the external load, the 98% quantile is selected as the characteristic value.

[0017] Further, the safety factor includes the safety factor of the load and the safety factor of the material of the tower, the safety factor of the load is the ratio of the load design value to the load characteristic value, and the safety factor of the material is the ratio of the material characteristic value to the material design value.

[0018] The advantages and beneficial effects of the present application are as follows:

[0019] The present application converts the traditional reliability optimization design problem into a deterministic optimization problem based on the safety factor, greatly reduces the calculation cost, and solves the problems of high test cost and difficulty in obtaining sample data according to the different reliability design requirements of the tower under different service environments. At the same time, combined with the safety factor checking strategy, the optimal design reliability under different reliabilities and arbitrary service environments is ensured, and the calculation cost and design redundancy are saved. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of the method in the embodiment of the present application.

[0021] Figure 2 is a wind turbine tower load action diagram in the embodiment of the present application.

[0022] Figure 3 is a safety factor checking result diagram in the embodiment of the present application. DETAILED DESCRIPTION

[0023] The specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0024] A wind turbine tower reliability optimization method based on a probability check safety factor is applied to a wind turbine tower, an advanced safety factor check strategy is adopted, a safety factor library under multiple reliabilities is constructed, the Wasserstein distance is used to measure environmental changes and guide safety factor selection, thereby constructing a deterministic optimization design problem based on a safety factor and solving it, as shown in Figure 1 , specifically including the following steps:

[0025] Step S1: For the tower type, determine the design variables, uncertainty variable distribution, and specify the target reliability interval.

[0026] Step S2: Combine the objective function and the constraint equation to build a multi-objective reliability optimization problem, solve it using a multi-objective algorithm to obtain MPP points (Most Probable Point) corresponding to different reliabilities; the multi-objective reliability optimization problem of the wind turbine tower is specifically:

[0027]

[0028] wherein, is the objective function of the optimization problem, and are the upper and lower bounds of the target reliability. is the reliability obtained based on the First order reliability method (FORM). The multi-objective algorithm uses the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with an elitist strategy.

[0029] Step S3: Based on each MPP, take each reliability corresponding MPP point as the design value of the uncertainty variable under the corresponding reliability, and construct the mapping relationship between each uncertainty variable design value and reliability.

[0030] Step S4: For each uncertainty variable distribution, select a fixed quantile as its characteristic value. For structure strength related uncertainty variables, to ensure design reliability, the 5% quantile is usually taken as the characteristic value, such as the compressive strength of materials. For external load related uncertainty variables, the 98% quantile is selected as the characteristic value.

[0031] According to the safety factor definition formula, obtain the safety factor library under different reliabilities ; the safety factor definition formula is:

[0032]

[0033]

[0034] in, For the load safety factor, It is the design load value. It is the load characteristic value; For material safety factor, These are characteristic values ​​of material properties. These are the design values ​​for material properties.

[0035] Step S5: In engineering applications, determine the combination of eigenvalues ​​and target reliability safety factors under the current environment;

[0036] For any environment, determine its load distribution parameters and select its fixed quantiles as the characteristic values ​​of the current environment. ;

[0037] Target reliability for design operating conditions The reliability offset required to calculate environmental load variations based on Wasserstein distance. The Wasserstein distance calculation formula is as follows:

[0038]

[0039] in, and These are the load distribution when checking the safety factor and the load distribution under the new environment, respectively. and This is the corresponding inverse cumulative distribution function. For a given load variable... (like: , , , Reliability offset The calculation method is as follows:

[0040]

[0041] in, Indicates the load characteristic value. In terms of target reliability The sensitivity at a given location is calculated using the following formula:

[0042]

[0043] in, It is the reliability of the corresponding target in the standard normal space. The most likely point, It is a constraint equation for gradient, The norm of the gradient.

[0044] Select the target reliability from the safety factor library. Corresponding safety factor combination .

[0045] Step S6: Construct an optimization design problem based on the safety factor method and solve it using the Sequential Quadratic Programming (SQP) algorithm to quickly obtain the optimal design value corresponding to the target reliability.

[0046] The optimization design problem based on the safety factor method has the following specific formula:

[0047]

[0048] Where g(·) represents the constraint equation incorporating the safety factor, Representing load variables (e.g.) The eigenvalues ​​(98th percentiles) of the new distribution (etc.). Indicates the characteristic value of material properties.

[0049] Example:

[0050] To verify the effectiveness of the method of this invention, this embodiment uses a 2MW offshore wind turbine tower from Generic as the design object. Using this wind turbine model, the parameters corresponding to the DLC1.4 operating condition (continuous gust intensity and wind direction change angle) were set in the BLADED software, and a ten-minute simulation was performed to extract the maximum tower load. Figure 2 As shown. Loads on each tower section (axial force) tangential bending moment tangential force axial bending moment The uncertainty characterization results and parameters are shown in Table 1.

[0051] Table 1. Uncertainty Characterization Results and Parameters for Each Tower Section Load

[0052]

[0053] In this case, the design objective is to minimize the amount of steel used in the tower. Using the same Q345 steel material, this objective is equivalent to minimizing the tower volume. The reliability constraint primarily considers axial buckling failure of the tower. The design variable is the tower wall thickness. (In other embodiments, it could also be the tower diameter, height, etc.). According to standard requirements, the tower buckling failure mode can be expressed as:

[0054]

[0055]

[0056] in, This represents the axial buckling failure function of the tower. This represents the tangential buckling failure function of the tower. and These are the axial and tangential stress resistances, respectively. and These are axial and tangential stresses, defined as follows:

[0057]

[0058]

[0059]

[0060]

[0061] in, and These are the buckling reduction coefficients for axial and tangential stresses, respectively; It is the buckling strength of the material; It is an uncertainty factor in aeroelasticity; It is the experimental uncertainty factor; It is an aerodynamic uncertainty factor; These are structural uncertainty factors, and the parameters of the above uncertainty variables are shown in Table 2.

[0062] Table 2 Uncertainty Variable Parameter Table for Examples

[0063]

[0064] The fourth tower section is selected as the design object. A multi-objective reliability optimization problem is constructed, as shown in the following equation. For this wind turbine, the target reliability interval is set to [2,4]. In the reliability analysis, the reliability... There is a one-to-one correspondence between the system failure probabilities and the interval [2,4], which represents the corresponding failure probability. In the NSGA-II algorithm, the population size for each generation is set to 100, and the maximum number of iterations is 100.

[0065]

[0066] Each solution corresponds to an MPP point The analysis revealed that the axial buckling constraint of the tower was an active constraint, while the tangential buckling failure was an inactive constraint. Therefore, only the axial buckling constraint was considered in subsequent design steps. , and distribution parameters, determine its eigenvalue , , , combined with the safety factor calculation formula, the safety factor library of the target reliability interval [2, 4] is obtained, as shown in Figure 3 In this case, based on 100 Pareto solution sets, 100 safety factor combinations can be obtained. If the target reliability is 3, the safety factor combination is , represents the safety factor of , and represents the safety factor of , both of which belong to the load safety factor represents the material safety factor.

[0067] In the face of different reliability requirements and different external environments, corresponding safety factor combinations can be selected to construct targeted optimization design problems. For example, if the external environment changes, the axial bending moment load obeys the Weibull distribution, and the distribution parameters are , the axial bending moment eigenvalue under this environment can be calculated as N. Therefore, the following optimization design model can be constructed:

[0068]

[0069] where g(·) represents the constraint equation obtained by incorporating the safety factor into the failure constraint equation, which is equivalent to and combination.

[0070] For different target reliabilities, select the corresponding safety factor combination in the safety factor library and variable eigenvalue, and solve the above equation to obtain the optimal design that meets the reliability requirements. Table 3 is the test result under the axial bending moment load obeys the Weibull distribution, and the distribution parameters are , which also proves the feasibility of the proposed method. Compared with the safety factor method based on sensitive factors (referred to as SF, Safety Factor) and the performance measurement method (PMA, Performance measurement approach), the results show that the multi-objective reliability safety factor library constructed by this method can be applied to different service environments and different reliability optimization designs in actual engineering, saving calculation time, reducing the number of performance function evaluations, and conforming to the usage habits of existing engineering applications.

[0071] Table 3 Tower design results for different target reliabilities for examples

[0072]

[0073] The above examples are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application has been described in detail with reference to the foregoing examples, those ordinarily skilled in the art should understand: the technical solutions recorded in the foregoing examples can still be modified, or some or all of the technical features can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A wind turbine tower reliability optimization method based on a probability check safety factor, characterized in that: structure parameters of the tower are taken as determined design variables, uncertain factors to which the tower is subjected are taken as uncertain variables, and a target reliability interval of an optimization target is determined according to an interval of a tower failure probability; based on a structure optimization target of the tower, a maximum possible point of the uncertain variables corresponding to a reliability is obtained within the target reliability interval; a design value of the uncertain variables under the corresponding reliability of the maximum possible point is selected, and a fixed quantile is selected as a characteristic value according to a distribution of the uncertain variables; safety factors under different reliabilities are obtained through the design value and the characteristic value; a reliability offset required by a change in the uncertain variables is calculated for a target reliability and a load distribution under different working conditions, and a corresponding safety factor is selected in combination with the reliability offset and the target reliability and integrated into the structure optimization of the tower to generate the design variables corresponding to the optimal tower under any load distribution and different reliabilities after adjustment.

2. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 1, characterized in that: a tower buckling failure mode is obtained according to a stress to which the tower is subjected and a stress resistance, a failure constraint is constructed through the tower buckling failure mode, a target reliability optimization is performed based on the failure constraint and the optimization target of the tower, and a maximum possible point of the uncertain variables corresponding to different reliabilities within a target reliability interval is obtained.

3. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 2, characterized in that: a fixed quantile is selected as a current environmental characteristic value according to a distribution of the uncertain variables determined by the environment; a weight coefficient is obtained by measuring a distance between a load distribution when the safety factor is checked and a load distribution under a new environment, the weight coefficient is multiplied by a sensitivity at the target reliability, and a ratio of a product to a current environmental characteristic value is a reliability offset required by a change in the environmental load.

4. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 3, characterized in that: The sensitivity is obtained by obtaining a most probable point corresponding to the target reliability in a standard normal space, and a square of a load partial derivative of the failure constraint at the most probable point is divided by a square of a norm of a gradient of the failure constraint at the most probable point.

5. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 3, characterized in that: The failure constraint is adjusted based on a characteristic value of the load variable under a new distribution, a material characteristic characteristic value, and a safety factor corresponding to the adjusted target reliability.

6. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 2, characterized in that: The uncertain variables include a material buckling strength, an uncertainty factor, and a load, and the load includes a force and a bending moment to which the tower is subjected.

7. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 6, characterized in that: The design variables are thicknesses of the tower, the optimization target is a minimum volume of the tower, the tower buckling failure mode is a difference between a stress resistance and a stress, the stress resistance is obtained by a product of a stress buckling reduction coefficient and the material buckling strength, the stress is obtained by a product of the force to which the tower is subjected, a product of a tower circumference and the thickness, a product of the bending moment, a product of a tower area and the thickness, and the uncertainty factor, and the sum of the two products.

8. The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 7, characterized in that: The safety factors include a safety factor of the material buckling strength, a safety factor of the force to which the tower is subjected, and a safety factor of the bending moment; and the safety factor integrated into the structure optimization of the tower is obtained by multiplying the material buckling strength, the force to which the tower is subjected, and the bending moment in the tower buckling failure mode by the respective safety factors. 9.The wind turbine tower reliability optimization method based on probability checking safety factor according to claim 1, characterized in that: For structure strength-related uncertain variables, a low percentage quantile is selected as a characteristic value. For the external load related uncertainty variables, high percentile quantiles are selected as characteristic values.

10. The method for optimizing the reliability of a wind turbine tower based on a probabilistic check of the safety factor according to claim 1, characterized in that: The safety factor includes a load safety factor and a material safety factor of the tower, the load safety factor being a ratio of a load design value to a load characteristic value, and the material safety factor being a ratio of a material characteristic characteristic value to a material characteristic design value.

Images