Offshore wind turbine support structure optimization design method and system based on agent model

Abstract

The application discloses a method and system for offshore wind turbine support structure optimization design based on a proxy model. The method is based on a Kriging proxy model to establish the relationship between the design parameters of the support structure of different capacity sizes of offshore wind turbines and the structural performance response of the optimization design, and simultaneously optimizes the design of the tower and the single pile foundation to find the optimization design with the lightest weight of the overall support structure. The method can avoid the problem of falling into a local optimal solution during the optimization design of the offshore wind turbine support structure, and can find only the lightest design of the tower or the single pile. The method and process used by the application can simultaneously check and optimize the design of the tower and the foundation structure after the load calculation, can find the global optimal design, can make the overall structure stronger, the stiffness distribution more uniform, and the stress more reasonable, and finally can achieve the lightest weight of the overall support structure. The optimization design based on the proxy model can reduce the number of computer simulation structure analysis, and can improve the optimization efficiency.

Description

Technical Field

[0001] This invention belongs to the field of offshore wind turbine support structure design technology, specifically involving an optimization design method and system for offshore wind turbine support structures based on a proxy model. Background Technology

[0002] Benefiting from technological advancements and increased scale in wind power, the prices of wind turbine units, wind power development investment costs, and operation and maintenance costs are showing a continuous downward trend. Looking at wind turbine unit prices, the support structure of offshore wind turbines includes two parts: the tower and the foundation. The cost of the turbine tower accounts for approximately 8% of the investment cost of an offshore wind power project, while the foundation, which mainly includes different foundation types such as monopiles, jacket foundations, and high-pile caps, generally accounts for about 14% of the investment cost of an offshore wind power project. In other words, the overall support structure cost accounts for approximately 22% of the total construction cost. Therefore, reducing the cost of the offshore wind power support structure can effectively reduce the levelized cost of electricity (LCOE) of offshore wind power.

[0003] Currently, domestic offshore wind power projects typically employ a step-by-step iterative design approach during the bidding process. Generally, the wind turbine manufacturer provides the tower design and guarantees the tower's quantity, and the tower weight is ranked and scored during the evaluation process. In the post-bid detailed design phase, the wind turbine manufacturer and the design institute sequentially optimize the tower and foundation designs, respectively. Under this process, the wind turbine manufacturer aims to provide the lightest locally optimal tower design, but this lightest tower design is often not the globally optimal design for the overall support structure. Existing research often only optimizes a portion of the overall support structure (tower or foundation), thus the results have certain limitations.

[0004] The design of offshore wind turbine support structures includes three parts: load calculation, tower design, and foundation design.

[0005] 1) Load calculation

[0006] Offshore wind turbine support structures are subjected to the combined effects of various environmental loads, including wind, waves, and currents. Most wind turbine manufacturers in the industry use GH-Bladed for integrated modeling and load calculation.

[0007] Integrated modeling comprises two aspects: inputting environmental conditions and building an overall support structure model. Environmental conditions include wind resource parameters, marine hydrological parameters, engineering geological parameters, and other special operating conditions (sea ice, earthquakes, typhoons, etc.). The overall support structure model includes the nose cone, tower, above-the-mud structure, and foundation (which can also be collectively referred to as the foundation structure).

[0008] The load calculation takes into account the anisotropic effects of wind and waves. According to IEC standards, it is necessary to consider various operating conditions such as normal power generation, emergency shutdown, start-up, normal shutdown, idling, and maintenance. Based on the combined distribution of wind and waves, it can be divided into more than 20,000 operating conditions.

[0009] 2) Tower Design

[0010] In tower design, it is necessary to verify the ultimate strength, buckling strength, and fatigue strength of the main tower structure and its local structures. Ultimate strength verification includes verification of local structures such as the tower body, tower flanges, portals and cable holes, and anchor cages; buckling strength verification includes verification of structures such as the tower body, portals and cable holes; fatigue strength verification includes verification of structures such as tower body welds, flange connection bolts, portal frames and cable holes, top flanges, and anchor cages.

[0011] 3) Basic Design

[0012] The main structural design includes strength and bearing capacity analysis under extreme sea states, normal service condition analysis, ship collision analysis, and seismic condition analysis. Load combinations consider the extreme combinations of waves, currents, and wind turbine operating loads at the most unfavorable water levels. Fatigue strength analysis utilizes the SN curve and Miner's linear cumulative damage theory for fatigue calculations. The cumulative damage degree of each pipe node under fatigue loads is calculated, and the cumulative damage degree is used to assess the fatigue resistance design safety of the structure.

[0013] Currently, most of the wind power industry in China adopts a step-by-step iterative design method. Figure 1 A schematic diagram of an integral support structure for a marine monopile foundation is provided. (For example...) Figure 1 As shown, the overall support structure is divided by the design interface, with the tower above the interface and the foundation structure below the interface. Figure 2 The flowchart of the step-by-step iterative design method is presented. First, the design institute provides the environmental input for the project. Based on this input, the wind turbine manufacturer provides the initial configuration of the tower and foundation, performs overall modeling and load calculations, and after obtaining the optimal tower, submits the loads at the design interface, tower configuration, and frequency requirements to the design institute. Next, the design institute checks and optimizes the foundation structure under the given loads and tower configuration, ensuring it meets the frequency requirements provided by the wind turbine manufacturer. Finally, the wind turbine manufacturer, after obtaining the optimized foundation structure, determines whether convergence has occurred. If convergence is achieved, the iteration ends; otherwise, the model is re-built and load calculations are performed. The convergence criteria here include two types: one is the design criterion for checking the tower and foundation according to specifications; the other is whether the difference in quality and frequency between the current and previous optimized designs is within 1%.

[0014] It's important to note that in current domestic offshore wind turbine support structure design methods, after determining the initial configuration (diameters of the tower and monopile foundations), 2-4 iterations are typically required for convergence. Each iteration necessitates load calculations and design optimization of the tower and foundation. Further optimizing the tower and monopile diameters to find the lightest overall support structure design is extremely time-consuming, impacting project schedules. Therefore, in actual engineering projects, to expedite the provision of construction drawings for the tower and monopile foundations, insufficient time is often allocated for optimization. Furthermore, the design and optimization of the tower and foundation are performed sequentially within two independent design domains, each aiming to find the optimal design within its respective domain. Consequently, in domestic projects, the final design often represents a locally optimal design for the lightest tower, rather than a globally optimal design for the lightest overall support structure.

[0015] If sensitivity-based optimization algorithms are used to solve for the optimal design of offshore wind turbine support structures, the sensitivity of structural performance to design variables can only be obtained through the finite difference method, which is extremely labor-intensive and cannot meet the requirements of actual construction schedules. Therefore, reducing the number of computer simulations and structural analyses required and improving optimization efficiency is one way to overcome this difficulty; another approach has also received widespread attention. This approach is to treat commercial structural analysis software as a black box, using it to calculate the performance response of complex structures on a batch of sample designs, and then constructing a surrogate model for the structural analysis model based on this calculation.

[0016] Depending on the surrogate model constructed, the main surrogate models include the Response Surface Method (RSM), Radial Basis Function (RBF), Support Vector Regression (SVR), and Kriging. In this invention, we recommend using the Kriging surrogate model. It should be noted that other surrogate model methods can also achieve the optimization objective.

[0017] The Kriging surrogate model can provide predicted values ​​and variances for sample points, making it suitable for highly nonlinear problems and capable of smoothly approximating the original model with fewer sample points. However, because the Kriging method is an interpolation method, it is relatively sensitive to numerical noise.

[0018] The Kriging model can be written as:

[0019] y(x)=F(β,x)+z(x)=f T β+z(x) (1)

[0020] Where β is the regression coefficient; f(x) is an approximation function, expressed as a 0th, first, or second-order polynomial of x; z(x) is the error of a random distribution, with the following statistical properties:

[0021] E[z(x)]=0

[0022]

[0023]

[0024] Where x i ,x j R(θ,x) is any two sample points in the sample set. i ,x j ) is a correlation function with parameter θ, which can be given as:

[0025]

[0026] Where n v It is the number of design variables. These are sample points x i ,x j The k-th component. Where R k (θ k ,d k There are many forms, including:

[0027] linear function R k (θ k ,d k )=max{0,1-θ k d k}

[0028] Quadratic polynomial function

[0029] Exponential function R k (θ k ,d k )=exp(-θ k d k )

[0030] Gaussian function

[0031] The Gaussian correlation function is the most widely used correlation function due to its good computational performance.

[0032] It is worth noting that we can construct the response surface with the minimum error by choosing different regression and correlation functions, which will be discussed further in Chapter 4.

[0033] Suppose there are n sample points, given a sample set S = [x1, x2, ..., xn].ns ] and its response set Y = [y1, y2, ..., y ns For model (1), for any point x to be measured new The predicted response value is:

[0034]

[0035] The prediction error is:

[0036]

[0037] in To ensure the unbiasedness of the simulation process, the mean of the error should be zero, that is:

[0038]

[0039] We can obtain:

[0040] F T cf = 0 (7)

[0041] The prediction variance of equation (4) is:

[0042]

[0043] in:

[0044]

[0045] This formula represents the new sample point x. new Spatial correlation with each sample point. At this point, the difference coefficient c can be determined by minimizing the prediction variance of the predicted values; the optimization model is as follows:

[0046]

[0047] By solving, we can obtain:

[0048]

[0049]

[0050] Substituting the c obtained from the above formula into equations (4) and (8), we obtain the point x to be measured. new Predicted values ​​and predicted value variance:

[0051]

[0052] From the regression problem The generalized least squares estimate can be obtained as follows:

[0053] β * =(F T R-1 F) -1 F T R -1 Y (13)

[0054] Substituting equation (13) into equation (12), and considering the residual expression Rγ * =Y-Fβ * We can obtain:

[0055]

[0056] Since matrices F, R, and vector Y are computed from given samples S, and x new They are irrelevant, so β ​​and γ are unrelated to x. new If it is irrelevant, then in equation (14) only the vector f(x) is present. new ) and r(x new ) and x new That is, for any point x to be measured... new We only need to find f(x) new ) and r(x new By doing so, we can obtain the predicted response value for this point.

[0057] Substituting equation (11) into equation (12), the prediction variance of the predicted value is:

[0058]

[0059] in:

[0060]

[0061] Therefore, equations (14) and (15) can be used to calculate the predicted value and the predicted variance of any point. These two equations contain two parameters. And θ. These two parameters can be calculated by maximizing the likelihood estimate of the response value. If z(x) follows a normal distribution, then y(x) should also follow a normal distribution, and its likelihood function is:

[0062]

[0063] Taking the logarithm of the above expression and removing the constant term, we get:

[0064]

[0065] Adjust the above equation Taking the partial derivative and setting it to zero, we get:

[0066]

[0067] Substituting (17) into (16) and ignoring the constant term, we get:

[0068]

[0069] The value of θ can be obtained by maximizing (17), thus solving the maximization problem:

[0070]

[0071] In summary, the construction of the Kriging surrogate model is transformed into a nonlinear unconstrained optimization problem.

[0072] If a sensitivity-based optimization algorithm is to be used to solve the optimal design of the offshore wind turbine support structure, the sensitivity of the structural performance to the design variables can only be obtained through the finite difference method, which is very labor-intensive and cannot meet the requirements of the actual construction period. Summary of the Invention

[0073] To address the problems existing in the prior art, this invention provides an optimization design method and system for offshore wind turbine support structures based on a proxy model. It optimizes the design of both the tower and foundation structures, and designs the offshore wind turbine support structure from the perspective of finding the globally optimal design, thereby reducing the levelized cost of electricity (LCOE) and design cycle of offshore wind power.

[0074] To achieve the above objectives, the technical solution adopted by this invention is: a method and system for optimizing the design of offshore wind turbine support structures based on a proxy model, comprising the following steps:

[0075] Based on the design variables of the determined wind turbine support structure, an initial sample set is generated within the design space using a sampling method;

[0076] Numerical simulation is performed on each sample point in the initial sample set to obtain the corresponding structural performance response results;

[0077] Select a surrogate model, and based on the initial sample set and structural performance response results, use the corresponding fitting or interpolation method to establish a mapping relationship between the initial sample set and structural performance response results;

[0078] Using a proxy model for structural optimization design, preliminary design results of the tower and foundation structures are obtained. During structural optimization design, the tower and foundation structures are checked and optimized simultaneously after load calculation.

[0079] The preliminary design results are simulated using numerical simulation to confirm their feasibility and optimality, and to verify whether they meet the convergence criteria. If the convergence criteria are not met, appropriate point addition criteria are selected to add sample points, the surrogate model is updated, and the updated surrogate model is used for optimization design until the convergence criteria are met, at which point the optimization iteration ends.

[0080] The design variables are the tower base diameter and the single pile diameter; the structural performance response is the tower mass, the single pile mass, and the overall support structure frequency.

[0081] An initial sample set was generated using a central composite experiment method.

[0082] The structural performance response consists of the tower mass, the mass of a single pile, and the frequency of the overall support structure.

[0083] The Kriging proxy model is adopted.

[0084] The convergence criterion is as follows:

[0085] Δ weigh / Weigh n-1 ≤1%

[0086] Δ frequency / Frequency n-1 ≤1%

[0087] Where, Δ weigh / Weigh n-1 ≤1% indicates whether the difference in tower mass and single pile mass between the optimized design obtained in this round and the previous round is within 1%; Δ frequency / Frequency n-1 ≤1% indicates whether the difference in the overall support structure frequency between the current and previous rounds of optimized design is within 1%.

[0088] The optimization formula in the process of using the surrogate model is as follows:

[0089] find: Diameter, wall thickness, and weld height of the tower base and monopile foundation.

[0090] minimum: overall supporting structure mass

[0091] subject to:

[0092] SRF ULS_tower ≥1 ①

[0093] SRF FLS_tower ≥1 ②

[0094] UC ULS_foundation ≤1 ③

[0095] Damage FLS_foundation ≤1 ④

[0096] The design compressive bearing capacity of a single pile foundation is less than or equal to the allowable compressive bearing capacity of a single pile foundation.

[0097] The design deformation value of a single pile foundation is less than or equal to the allowable deformation value of a single pile foundation.

[0098] Among them: SRF ULS_tower SRF indicates the safety margin of the tower structure under extreme conditions. FLS_tower Indicates the safety margin of the tower structure under fatigue conditions; UC ULS_foundation Damage represents the element index of a monopile foundation structure under ultimate limit state. FLS_foundation This indicates fatigue damage in a single-pile foundation structure.

[0099] ①The lightest design for the tower is not the lightest design for the overall support structure;

[0100] ② The tower diameter corresponding to the lightest overall support structure design is relatively large: 6.0m for 3-5MW units and 7.0m for 6-8MW units. The corresponding single pile diameter is slightly larger than the tower base diameter by 0.5m-1.0m, that is, the single pile diameter of 3-5MW units is 6.5-7.0m and the single pile diameter of 6-8MW units is 7.5-8.0m.

[0101] Another aspect of the present invention is to provide an optimization design system for offshore wind turbine support structures based on a proxy model, including an initial sample set generation module, a structural performance response acquisition module, a mapping relationship construction module, an optimization module, and an iterative verification module;

[0102] The initial sample set generation module generates an initial sample set within the design space based on the design variables of the wind turbine support structure using a sampling method.

[0103] The structural performance response acquisition module is used to perform numerical simulation on each sample point in the initial sample set to obtain the corresponding structural performance response results.

[0104] The mapping relationship construction module is used to select a proxy model and, based on the initial sample set and the structural performance response results, establish a mapping relationship between the initial sample set and the structural performance response results using the corresponding fitting or interpolation method.

[0105] The optimization module uses a proxy model to perform structural optimization design and obtains preliminary design results for the tower and foundation structures. During the structural optimization design, the tower and foundation structures are checked and optimized simultaneously after load calculation.

[0106] The iterative verification module simulates the obtained preliminary design results using numerical simulation to confirm their feasibility and optimality, and verifies whether the convergence criterion is met. If the convergence criterion is not met, a suitable point addition criterion is selected to add sample points, update the surrogate model, and use the updated surrogate model for optimization design until the convergence criterion is met, at which point the optimization iteration ends.

[0107] The present invention also provides a computer device, including a processor and a memory. The memory is used to store a computer executable program. The processor reads the computer executable program from the memory and executes it. When the processor executes the computer executable program, it can realize the offshore wind turbine support structure optimization design method based on the proxy model described in the present invention.

[0108] Compared with the prior art, the present invention has at least the following beneficial effects:

[0109] The method described in this invention avoids getting stuck in local optima during the optimization design of offshore wind turbine support structures, which would only result in finding the lightest design for the tower or monopile. Using the method and process of this invention, the tower and foundation structures are simultaneously checked and optimized after load calculation, leading to a globally optimal design. This results in a stronger overall structure, more uniform stiffness distribution, and more rational stress distribution, ultimately achieving the lightest overall support structure weight and reducing the cost of offshore wind turbine support structures. Optimization design based on a surrogate model reduces the number of computer simulations required for structural analysis, improving optimization efficiency. Attached Figure Description

[0110] The exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which will make the above and other features and advantages of the present invention more apparent. In the accompanying drawings:

[0111] Figure 1 This is a schematic diagram of the marine support structure.

[0112] Figure 2 This is a step-by-step iterative design method for offshore wind turbine support structures.

[0113] Figure 3 This is the process of optimizing the proxy model algorithm.

[0114] Figure 4 It is a process for the overall optimization design of offshore wind turbine support structures.

[0115] Figure 5 These are the initial sample points for the tower base diameter and monopile diameter of 3-5MW offshore wind turbine units.

[0116] Figure 6 These are the initial sample points for the tower base diameter and monopile diameter of 6-8MW offshore wind turbine units.

[0117] Figure 7 It is a quality proxy model for 3-5MW offshore wind turbine towers.

[0118] Figure 8 It is a quality proxy model for the overall support structure of 3-5MW offshore wind turbine units.

[0119] Figure 9It is a quality proxy model for 6-8MW offshore wind turbine towers.

[0120] Figure 10 It is a quality proxy model for the overall support structure of 6-8MW offshore wind turbines. Detailed Implementation

[0121] The following is combined with Figures 3-10 The present invention will be further described in detail with reference to specific embodiments.

[0122] refer to Figure 3 The general flowchart of the surrogate model optimization algorithm includes the following steps:

[0123] Step 1: Obtain the design characteristics of the offshore support structure through engineering project experience and screen the design variables. In this invention, the design variables are the tower base diameter and the single pile diameter.

[0124] Step 2: An initial sample set is generated within the design space using a sampling method. This invention employs a central composite experiment method.

[0125] Step 3: Perform numerical simulation analysis on each sample point in the initial sample set to obtain the corresponding structural performance response. In this invention, the structural performance response includes the tower mass, the mass of a single pile, and the frequency of the overall support structure.

[0126] Step 4: Select a surrogate model (Kriging surrogate model is recommended, but not limited to) and use the appropriate fitting or interpolation method to establish the mapping relationship between the inputs and outputs in Steps 2 and 3.

[0127] Step 5: Use the proxy model to perform structural optimization design and obtain the preliminary design results of the tower and foundation structure. When performing structural optimization design, check and optimize the tower and foundation structure simultaneously after load calculation.

[0128] Step 6: Perform numerical simulation analysis on the preliminary design results obtained in Step 5 to confirm their feasibility and optimality, and verify whether they meet the convergence criteria. If the convergence criteria are not met, select an appropriate point addition criterion to add sample points, update the surrogate model, and return to Step 5. If the convergence criteria are met, the optimization iteration ends. The convergence criteria here refer to whether the differences between the response values ​​of the optimized design obtained in this round and the previous round, as well as physical quantities such as tower mass, single pile mass, and overall support structure frequency, are within 1%.

[0129] Following the numerical simulation method for sample points in step 2 as described above, this invention proposes an overall optimization design method for offshore wind turbine support structures based on a surrogate model; (Reference) Figure 4The flowchart illustrates the overall optimization design method for offshore wind power support structures. The biggest difference between this method and the step-by-step iterative design method is that the overall optimization design method simultaneously checks and optimizes the tower and foundation structures after load calculation, seeking the globally optimal design with the lightest overall support structure within the entire design domain. Wind turbine manufacturers and design institutes, after obtaining the load data, simultaneously optimize the tower and foundation structures, designing from the perspective of finding the globally optimal design. Currently, the procurement cost of materials for offshore support structures primarily considers quality factors. The overall optimization design method can reduce the overall mass of the support structure, thus achieving the goal of reducing the levelized cost of electricity (LCOE) for offshore wind power. The optimization formula is as follows:

[0130] find: Diameter, wall thickness, and weld height of the tower base and monopile foundation.

[0131] minimum: overall supporting structure mass

[0132] subject to:

[0133] SRF ULS_tower ≥1 ①

[0134] SRF FLS_tower ≥1 ②

[0135] UC ULS_foundation ≤1 ③

[0136] Damage FLS_foundation ≤1 ④

[0137] The design compressive bearing capacity of a single pile foundation is less than or equal to the allowable compressive bearing capacity of a single pile foundation.

[0138] The design deformation value of a single pile foundation is less than or equal to the allowable deformation value of a single pile foundation.

[0139] Among them: SRF ULS_tower SRF indicates the safety margin of the tower structure under extreme conditions. FLS_tower Indicates the safety margin of the tower structure under fatigue conditions; UC ULS_foundation Damage represents the element index of a monopile foundation structure under ultimate limit state. FLS_foundation This indicates fatigue damage in a single-pile foundation structure.

[0140] The optimal support structure can be obtained by using the surrogate model-based optimization design method for offshore wind turbine support structures proposed in this invention:

[0141] ①The lightest design for the tower is not the lightest design for the overall support structure;

[0142] ② The tower diameter corresponding to the lightest overall support structure design is relatively large (6.0m for 3-5MW units and 7.0m for 6-8MW units), and the corresponding single pile diameter is slightly larger than the tower base diameter by 0.5m-1.0m, that is, 6.5-7.0m for 3-5MW units and 7.5-8.0m for 6-8MW units.

[0143] Taking the optimized design of the support structure for 3-5MW and 6-8MW offshore wind turbines as an example:

[0144] Step 1: Select the tower base diameter and single pile diameter as design variables.

[0145] Step 2, use the central composite assay method to generate initial sample points (e.g., Figure 5 These are the initial sample points for the tower base diameter and monopile diameter of 3-5MW offshore wind turbine units; Figure 6 These are the initial sample points for the tower base diameter and monopile diameter of 6-8MW offshore wind turbine units.

[0146] Step 3: Optimize the overall support structure of the offshore wind turbine to obtain the corresponding structural performance response (tower mass, monopile mass, overall support structure frequency, etc.).

[0147] Step 4: Select the Kriging surrogate model and, using the appropriate fitting or interpolation method, establish the mapping relationship between the inputs and outputs from steps 2) and 3) (e.g., ...). Figure 7 and Figure 8 It is a quality proxy model for the tower of a 3-5MW offshore wind turbine and a quality proxy model for the overall support structure; Figure 9 and Figure 10 These are the initial sample points for the tower base diameter and monopile diameter of 6-8MW offshore wind turbine units.

[0148] Step 5: Use a proxy model to replace the original model for structural optimization design.

[0149] Step 6: Verify the optimized design obtained in Step 5 using the original model to confirm its feasibility and optimality, and check the convergence criteria. If the criteria are not met, select an appropriate addition criterion to add sample points, update the surrogate model, and return to Step 5. If the criteria are met, the optimization iteration ends.

[0150] The convergence criterion here refers to whether the difference between the response values ​​(physical quantities such as tower mass, single pile mass, and overall support structure frequency) obtained from the current round and the previous round of optimization design is within 1%.

[0151] The optimized column is as follows:

[0152] find: tower base diameter, single pile diameter

[0153] minimum: overall supporting structure mass

[0154] subject to:

[0155] SRF ULS_tower ≥1

[0156] SRF FLS_tower ≥1

[0157] UC ULS_foundation ≤1

[0158] Damage FLS_foundation ≤1

[0159] The design compressive bearing capacity of a single pile foundation shall be less than or equal to the allowable compressive bearing capacity of the single pile foundation.

[0160] The design deformation value of a single pile foundation is less than or equal to the allowable deformation value of a single pile foundation.

[0161] Table 1 Calculation results of support structure for 3-5MW offshore wind turbines

[0162]

[0163] Table 2 Calculation results of support structure for 6-8MW offshore wind turbines

[0164]

[0165]

[0166] Tables 1 and 2 present the calculation results of the support structures for 3-5MW and 6-8MW offshore wind turbines based on the surrogate model optimization algorithm, respectively. For the 3-5MW turbine, the lightest tower design is Dtower = 5.5m, Dmonopil = 6.0m; the lightest overall support structure design is Dtower = 6.0m, Dpil = 6.5m. For the 6-8MW turbine, the lightest tower design is Dtower = 6.5m, Dmonopil = 8.5m; the lightest overall support structure design is Dtower = 7.0m, Dpil = 7.5m.

[0167] The characteristics of the optimal support structure that can be obtained through this example are:

[0168] 1) The lightest design for the tower is not the lightest design for the overall support structure;

[0169] 2) The tower diameter corresponding to the lightest overall support structure design is larger (6.0m for 3-5MW units and 7.0m for 6-8MW units), and the diameter of the corresponding single pile is 0.5m larger than the diameter of the tower base, i.e. 6.5m for 3-5MW units and 7.5m for 6-8MW units.

[0170] The offshore wind turbine support structure optimization design system based on the surrogate model includes an initial sample set generation module, a structural performance response acquisition module, a mapping relationship construction module, an optimization module, and an iterative verification module.

[0171] The initial sample set generation module generates an initial sample set within the design space based on the design variables of the wind turbine support structure using a sampling method.

[0172] The structural performance response acquisition module is used to perform numerical simulation on each sample point in the initial sample set to obtain the corresponding structural performance response results.

[0173] The mapping relationship construction module is used to select a proxy model and, based on the initial sample set and the structural performance response results, establish a mapping relationship between the initial sample set and the structural performance response results using the corresponding fitting or interpolation method.

[0174] The optimization module uses a proxy model to perform structural optimization design and obtains preliminary design results for the tower and foundation structures. During the structural optimization design, the tower and foundation structures are checked and optimized simultaneously after load calculation.

[0175] The iterative verification module simulates the obtained preliminary design results using numerical simulation to confirm their feasibility and optimality, and verifies whether the convergence criterion is met. If the convergence criterion is not met, a suitable point addition criterion is selected to add sample points, update the surrogate model, and use the updated surrogate model for optimization design until the convergence criterion is met, at which point the optimization iteration ends.

[0176] The present invention also provides a computer device, including a processor and a memory. The memory is used to store a computer executable program. The processor reads the executable program from the memory and executes it. When the processor executes the computer executable program, it can realize the offshore wind turbine support structure optimization design method based on the proxy model described in the present invention.

[0177] The computer device may be a laptop, tablet, desktop computer, or workstation.

[0178] The processor can be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or an off-the-shelf programmable gate array (FPGA).

[0179] The memory described in this invention can be an internal storage unit of a laptop, tablet, desktop computer, mobile phone, or workstation, such as memory or hard disk; or it can be an external storage unit, such as a portable hard disk or flash memory card.

Claims

1. A method for optimizing the design of offshore wind turbine support structures based on a surrogate model, characterized in that, Includes the following steps: Based on the design variables of the determined wind turbine support structure, an initial sample set is generated within the design space using a sampling method; Numerical simulation is performed on each sample point in the initial sample set to obtain the corresponding structural performance response results; Select a surrogate model, and based on the initial sample set and structural performance response results, use the corresponding fitting or interpolation method to establish a mapping relationship between the initial sample set and structural performance response results; Using a proxy model for structural optimization design, preliminary design results of the tower and foundation structures are obtained. During structural optimization design, the tower and foundation structures are checked and optimized simultaneously after load calculation. The preliminary design results were simulated using numerical simulation to confirm their feasibility and optimality, and to verify whether they met the convergence criteria. If the convergence criterion is not met, select an appropriate addition criterion to add sample points, update the surrogate model, and use the updated surrogate model for optimization design until the convergence criterion is met, then the optimization iteration ends. The optimization formula in the process of using the surrogate model is as follows: Find: Diameter, wall thickness, and weld height of the tower base and monopile foundation. minimum: overall supporting structure mass subject to: SRF ULS_tower ≥1① SRF FLS_tower ≥1② UC ULS_foundation ≤1③ Damage FLS_foundation ≤1④ The design compressive bearing capacity of a single pile foundation is less than or equal to the allowable compressive bearing capacity of a single pile foundation. The design deformation value of a single pile foundation is less than or equal to the allowable deformation value of a single pile foundation. in: SRF ULS_tower This indicates the safety margin of the tower structure under extreme conditions; SRF FLS_tower This indicates the safety margin of the tower structure under fatigue conditions; UC ULS_foundation The unit index represents the single pile foundation structure under the ultimate limit state; Damage FLS_foundation This indicates fatigue damage in a single-pile foundation structure.

2. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, The design variables are the tower base diameter and the single pile diameter; the structural performance response is the tower mass, the single pile mass, and the overall support structure frequency.

3. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, An initial sample set was generated using a central composite experiment method.

4. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, The structural performance response consists of the tower mass, the mass of a single pile, and the frequency of the overall support structure.

5. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, The Kriging proxy model is adopted.

6. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, The convergence criterion is as follows: in, This indicates whether the difference in tower mass and single pile mass between the optimized designs obtained in this round and the previous round is within 1%; This indicates whether the frequency difference of the overall support structure obtained from the optimized design in this round and the previous round is within 1%.

7. The method for optimizing the design of offshore wind turbine support structures based on a surrogate model according to claim 1, characterized in that, ①The lightest design for the tower is not the lightest design for the overall support structure; ② The tower diameter corresponding to the lightest overall support structure design is relatively large: 6.0m for 3-5MW units and 7.0m for 6-8MW units. The corresponding single pile diameter is slightly larger than the tower base diameter by 0.5m-1.0m, that is, the single pile diameter of 3-5MW units is 6.5-7.0m and the single pile diameter of 6-8MW units is 7.5-8.0m.

8. A system for optimizing the design of offshore wind turbine support structures based on a surrogate model, characterized in that, It includes an initial sample set generation module, a structural performance response acquisition module, a mapping relationship construction module, an optimization module, and an iterative verification module; The initial sample set generation module generates an initial sample set within the design space based on the design variables of the wind turbine support structure using a sampling method. The structural performance response acquisition module is used to perform numerical simulation on each sample point in the initial sample set to obtain the corresponding structural performance response results. The mapping relationship construction module is used to select a proxy model and, based on the initial sample set and the structural performance response results, establish a mapping relationship between the initial sample set and the structural performance response results using the corresponding fitting or interpolation method. The optimization module uses a proxy model to perform structural optimization design and obtains preliminary design results for the tower and foundation structures. During the structural optimization design, the tower and foundation structures are checked and optimized simultaneously after load calculation. The iterative verification module uses numerical simulation to simulate the obtained preliminary design results, confirm their feasibility and optimality, and verify whether they meet the convergence criteria. If the convergence criterion is not met, a suitable addition criterion is selected to add sample points, the surrogate model is updated, and the updated surrogate model is used for optimization design until the convergence criterion is met, at which point the optimization iteration ends; the optimization formula in the surrogate model optimization process is as follows: Find: Diameter, wall thickness, and weld height of the tower base and monopile foundation. minimum: overall supporting structure mass subject to: SRF ULS_tower ≥1① SRF FLS_tower ≥1② UC ULS_foundation ≤1③ Damage FLS_foundation ≤1④ The design compressive bearing capacity of a single pile foundation is less than or equal to the allowable compressive bearing capacity of a single pile foundation. The design deformation value of a single pile foundation is less than or equal to the allowable deformation value of a single pile foundation. in: SRF ULS_tower This indicates the safety margin of the tower structure under extreme conditions; SRF FLS_tower This indicates the safety margin of the tower structure under fatigue conditions; UC ULS_foundation The unit index represents the single pile foundation structure under the ultimate limit state; Damage FLS_foundation This indicates fatigue damage in a single-pile foundation structure.

9. A computer device, characterized in that, It includes a processor and a memory, the memory being used to store a computer-executable program, the processor reading the computer-executable program from the memory and executing it, and the processor executing the computer-executable program being able to implement the offshore wind turbine support structure optimization design method based on the proxy model as described in any one of claims 1 to 7.

Images