An offshore wind turbine foundation scouring intelligent early warning method and device and a storage medium

Abstract

This invention discloses an intelligent early warning method, device, and storage medium for offshore wind turbine foundation scour, belonging to the field of structural safety early warning technology. The method includes the following steps: obtaining the foundation structural parameters of the offshore wind turbine and establishing a finite element model; performing foundation scour simulation and collecting time-domain data of pile top displacement and pile nodal stress; establishing several surrogate models, dividing the time-domain data into training and validation sets, and obtaining training and validation results; using two validation indices, validation error and leave-one-out error, to select the best surrogate model; using different scour conditions as input parameters to obtain the wind turbine foundation response, comparing it with a set safety early warning threshold, and issuing a safety alarm if the wind turbine foundation response reaches or exceeds the safety early warning threshold. This invention can solve the problems of traditional experimental methods, such as the difficulty in reproducing initial conditions, the influence of experimental scale on measurement results, and high costs, as well as the shortcomings of finite element simulation in achieving full coverage of scour conditions.

Description

Technical Field

[0001] This invention relates to an intelligent early warning method, device, and storage medium for offshore wind turbine foundation scour, belonging to the field of engineering testing technology. Background Technology

[0002] With the increasing global demand for clean energy, offshore wind energy, as a clean and renewable energy source, is receiving more and more attention. However, the construction and operation of offshore wind farms face many technical challenges, among which the problem of foundation scour is particularly prominent. Seabed scour not only causes changes in the natural frequency of the turbine support structure but may also reduce the buckling capacity of the structure, seriously threatening the stable operation and economic benefits of offshore wind farms.

[0003] Existing research on the scour characteristics of offshore wind turbine foundations mainly relies on finite element simulation and experiments. However, finite element simulation requires a large amount of computation, and simulating the foundation response under different scour effects consumes significant computational resources, making it impossible to comprehensively cover complex and varying scour conditions. Experimental methods, on the other hand, suffer from drawbacks such as difficulty in reproducing initial conditions, the influence of experimental scale on measurement results, and high experimental costs. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide an intelligent early warning method, device and storage medium for scour of offshore wind turbine foundations. This invention can solve the problems of the difficulty in reproducing the initial conditions of traditional experimental methods, the influence of experimental scale on measurement results, and high cost, as well as the shortcomings of finite element simulation in achieving full coverage of scour conditions.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A method for intelligent early warning of scour of offshore wind turbine foundations includes the following steps:

[0007] Obtain the basic structural parameters of the offshore wind turbine and establish a finite element model based on the basic structural parameters;

[0008] The established finite element model was used to simulate the foundation scour, and time-domain data of pile top displacement and pile nodal stress were collected.

[0009] Several proxy models are established, and the time-domain data is divided into training set and validation set. The data is then imported into each proxy model for training and validation, and the training and validation results are obtained.

[0010] Based on the training and validation results, using two validation metrics, validation error and leave-one-out error, the surrogate model that meets the prediction accuracy and requires the smallest training set is selected as the best surrogate model.

[0011] Using different scouring conditions as input parameters for the optimal surrogate model, the wind turbine foundation response predicted by the optimal surrogate model is obtained, and the wind turbine foundation response is compared with the set safety warning threshold. If the wind turbine foundation response reaches or exceeds the safety warning threshold, a safety alarm is triggered.

[0012] The process of obtaining the basic structural parameters of the offshore wind turbine and establishing a finite element model based on the basic structural parameters includes: establishing a corresponding finite element model using ABAQUS software based on the offshore wind turbine foundation structural drawings.

[0013] The process of simulating foundation scour using the established finite element model and collecting time-domain data on pile top displacement and pile nodal stress includes: using the Coulomb friction model to simulate pile-soil interaction; using the birth and death element analysis method to remove elements in a specified area to simulate the local scour process; obtaining the wind turbine foundation stress cloud map through static analysis; and then obtaining the time-domain data on pile top displacement and pile nodal stress through finite element post-processing of the finite element model.

[0014] The proxy model is built based on the Uqlab toolbox in Matlab software, including the Kriging model, PCE (polynomial chaos expansion) model, PCK (polynomial chaos Kriging) model and SVR (support vector machine regression) model.

[0015] The input parameters of the surrogate model are a one-dimensional vector {X1,X2,X3,…}, including but not limited to the ratio of scour pit depth to pile diameter, the ratio of scour pit diameter to pile diameter, and the scour pit angle. The output parameters are a one-dimensional vector {Y1,Y2,Y3,…}, including but not limited to the foundation pile top displacement and pile joint stress. The input and output parameters are combined accordingly and randomly divided into a training set N. tra and validation set N val Two parts.

[0016] Verification error E val And leave one error E Loo The calculation methods are as follows:

[0017]

[0018]

[0019] In the formula: Let As a verification set K is the sample mean of the responses in the validation set, and K is the number of samples in the validation set. To obtain the surrogate model after removing the i-th sample point from the complete experimental design, It is the sample mean of the experimental design response. To verify the X value in the set, To verify the Y values ​​in the set, This is the output of the trained proxy model.

[0020] The standard for prediction accuracy is: E val ≤10 -2 And E Loo ≤10 -2 .

[0021] The warning threshold is specifically Δu = H / 75, where Δu is the horizontal displacement of the tall structure under wind-dominated load, and H is the height of the tall structure.

[0022] Secondly, the present invention provides an intelligent early warning device for offshore wind turbine foundation scour, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0023] Thirdly, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the steps of the intelligent early warning method for offshore wind turbine foundation scour.

[0024] The beneficial effects of this invention are as follows: This invention provides an intelligent early warning method, device, and storage medium for offshore wind turbine foundation scour. It uses a small amount of finite element calculation data to train four different surrogate models, and then uses two indicators to select the surrogate model with the best prediction effect. This enables accurate prediction of the response of offshore wind turbine foundations under different scour conditions. It overcomes the shortcomings of traditional experimental methods, such as the difficulty in reproducing initial conditions and high cost, and the inability of finite element simulation to achieve full coverage of scour conditions. It achieves "real-time" and "accurate" response prediction and can provide early warning, supervision, and judgment on structural safety and reliability. Attached Figure Description

[0025] Figure 1 The flowchart below shows a method for intelligent early warning of scour of offshore wind turbine foundations according to the present invention.

[0026] Figure 2 This is a finite element model diagram of the offshore wind turbine foundation established in this invention;

[0027] Figure 3 This is a comparison chart of the prediction accuracy of different proxy models in this invention;

[0028] Figure 4 This is a comparison chart of the errors of the displacement-stress prediction model established in this invention. Detailed Implementation

[0029] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solution of the present invention more clearly, and should not be used to limit the scope of protection of the present invention.

[0030] Example 1

[0031] This invention discloses an intelligent early warning method for offshore wind turbine foundation scour, comprising the following steps:

[0032] Step 1: Obtain the basic structural parameters of the offshore wind turbine and establish a finite element model based on the basic structural parameters;

[0033] Step 2: Perform a foundation scour simulation on the established finite element model and collect time-domain data of pile top displacement and pile nodal stress.

[0034] Step 3: Establish several proxy models, divide the time-domain data into training and validation sets, and import them into each proxy model for training and validation to obtain training and validation results.

[0035] Step 4: Based on the training and validation results, the best surrogate model is selected by using two validation metrics, validation error and leave-one-out error, which meet the prediction accuracy requirements and minimize the required training set.

[0036] Step 5: Using different scouring conditions as input parameters for the optimal surrogate model, obtain the wind turbine foundation response predicted by the optimal surrogate model, and compare the wind turbine foundation response with the set safety warning threshold. If the wind turbine foundation response reaches or exceeds the safety warning threshold, a safety alarm will be triggered.

[0037] This invention utilizes a small amount of finite element analysis data to train four different surrogate models. Then, using two metrics, it selects the surrogate model with the best predictive performance. This enables accurate prediction of the response of offshore wind turbine foundations under different scour conditions. It overcomes the shortcomings of traditional experimental methods, such as the difficulty in reproducing initial conditions and high costs, and the inability of finite element simulation to fully cover scour conditions. It achieves "real-time" and "accurate" response prediction and can provide early warning monitoring and judgment on structural safety and reliability.

[0038] Example 2

[0039] This invention discloses an intelligent early warning method for offshore wind turbine foundation scour, comprising the following steps:

[0040] Step one involves obtaining the foundation structural parameters of the offshore wind turbine and establishing a finite element model based on these parameters. The specific modeling method is as follows: Shell elements are used to simulate the foundation structure of a large-diameter monopile wind turbine, with dimensions based on a 7MW wind turbine: tower height 106.3m, tower diameter 5-7.7m, wall thickness 36mm, and embedment depth 35m. Solid elements are used to simulate the soil, creating a soil model with a diameter of 20D and a thickness of 50m. The interaction between the pile and soil is simulated using a Coulomb friction model. Figure 2 This is a finite element model diagram of the offshore wind turbine foundation established in this invention.

[0041] Step two involves simulating foundation scour using the established finite element model, collecting time-domain data on pile top displacement and pile joint stress. Specifically, the method involves using birth and death element analysis to simulate the localized scour process, obtaining a stress cloud map of the wind turbine foundation through static analysis, and then using finite element post-processing to obtain stress and displacement data for specified nodes.

[0042] Step 3: Establish several surrogate models, dividing the time-domain data into training and validation sets, and importing them into each surrogate model for training and validation, obtaining the training and validation results. Specifically, using the Uqlab toolbox in Matlab software, four surrogate models are built: Kriging model, PCE model, PCK model, and SVR model. The input parameters of the surrogate models are one-dimensional vectors {X1, X2, X3, ...}, including but not limited to the ratio of scour pit depth to pile diameter, the ratio of scour pit diameter to pile diameter, and scour pit angle, etc. The output parameters are one-dimensional vectors {Y1, Y2, Y3, ...}, including but not limited to foundation pile top displacement and pile joint stress, etc. The obtained input and output parameters are combined accordingly and randomly divided into training sets N. tra and validation set N val The study consists of two parts, using the training set data to train four agent models.

[0043] Step 4: Based on the training and validation results, the best surrogate model is selected by using two validation metrics: validation error and leave-one-out error. The surrogate model that meets the prediction accuracy requirements and minimizes the required training set is selected as the best surrogate model.

[0044] Verification error E val And leave one error E Loo The calculation methods are as follows:

[0045]

[0046] In the formula: Let As a verification set K is the sample mean of the responses in the validation set, and K is the number of samples in the validation set. To obtain the surrogate model after removing the i-th sample point from the complete experimental design, It is the sample mean of the experimental design response. To verify the X value in the set, To verify the Y values ​​in the set, This is the output of the trained proxy model.

[0047] The trained surrogate model achieves the required prediction accuracy as follows: E val ≤10 -2 And E Loo ≤10 -2For multiple surrogate models that meet this condition, compare their required training set sizes; the model with the smaller required training set is the optimal surrogate model for wind turbine foundation response prediction. For example... Figure 3 The graph compares the prediction accuracy of different surrogate models. The data comparison shows that the PCK surrogate model meets the accuracy requirements and is therefore selected as the best surrogate model. Figure 4 The graph shows the error comparison between the displacement-stress prediction model and the PCK surrogate model. It can be seen from the graph that the error of the PCK surrogate model in displacement-stress prediction is relatively small.

[0048] Step 5: Using different scour conditions as input parameters for the optimal surrogate model, obtain the wind turbine foundation response predicted by the optimal surrogate model, and compare the wind turbine foundation response with the set safety warning threshold. If the wind turbine foundation response reaches or exceeds the safety warning threshold, a safety alarm is triggered. Specifically, the warning threshold is determined according to GB50135: "Code for Design of Tall Structures," which stipulates that when a tall structure is under wind-dominant loads, its horizontal displacement shall not exceed 1 / 75 of its height. Therefore, Δu = H / 75 = 950 mm is defined as the tower top displacement warning limit value, and the wind turbine foundation node stress under this state is defined as the pile stress warning limit value.

[0049] Example 3

[0050] This embodiment discloses an intelligent early warning device for offshore wind turbine foundation scour, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method described in Embodiment 1 or 2.

[0051] Example 4

[0052] This embodiment describes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the method described in Embodiment 1 or 2.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0056] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0057] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An offshore wind turbine foundation scour intelligent early warning method, characterized in that: The method comprises the following steps: obtaining foundation structure parameters of the offshore wind turbine, and establishing a finite element model according to the foundation structure parameters; performing foundation scour simulation on the established finite element model, collecting time-domain data of pile top displacement and pile body node stress, specifically: performing pile-soil interaction simulation by using the Coulomb friction model, removing elements in a specified area to simulate the process of local scour by using the birth-death element analysis method, obtaining a wind turbine foundation stress cloud through static analysis, and then obtaining the time-domain data of the pile top displacement and the pile body node stress through finite element post-processing of the finite element model; A plurality of proxy models are established, and the time domain data is divided into a training set and a verification set, which are respectively introduced into each proxy model for training and verification to obtain training and verification results, wherein the input parameters of the proxy model are a one-dimensional vector {X1, X2, X3, …}, including but not limited to the ratio of the scour pit depth to the pile diameter, the ratio of the scour pit diameter to the pile diameter and the scour pit angle, and the output parameters are a one-dimensional vector {Y1, Y2, Y3, …}, including but not limited to the foundation pile top displacement and the pile body node stress; the input parameters and the output parameters are correspondingly combined and randomly divided into a training set N tra and a verification set N val two parts; based on the training and verification results, using two verification indexes of verification error and leave-one-out error, optimizing the proxy model that meets the prediction accuracy and requires the smallest training set as the best proxy model; taking different scour conditions as input parameters of the best proxy model, obtaining the wind turbine foundation response predicted by the best proxy model, and comparing the wind turbine foundation response with the set safety warning threshold, if the wind turbine foundation response reaches or exceeds the safety warning threshold, safety warning is performed.

2. The offshore windmill foundation scouring intelligent warning method according to claim 1, characterized in that: The obtaining of the foundation structure parameters of the offshore wind turbine and the establishment of the finite element model according to the foundation structure parameters comprises: establishing a corresponding finite element model by using ABAQUS software according to the offshore wind turbine foundation structure drawings.

3. The offshore windmill foundation scouring intelligent warning method according to claim 1, characterized in that: The proxy model is constructed based on the Uqlab toolbox in Matlab software, and comprises a Kriging model, a PCE model, a PCK model and a SVR model.

4. The offshore windmill foundation scouring intelligent warning method according to claim 1, characterized in that: The verification error E val and the leave-one-out error E Loo The calculation methods are respectively: wherein: let As the validation set, is the sample mean of the validation set response, K is the number of samples in the validation set, is the surrogate model after removing the ith sample point from the full experimental design, is the sample mean of the experimental design response, is the X value in the validation set, is the Y value in the validation set, is the output of the trained surrogate model.

5. A method of intelligent scour warning for offshore windmill foundations according to claim 4, characterized in that: The standard for the prediction accuracy is: E val ≤ 10 -2 and E Loo ≤ 10 -2 .

6. The offshore windmill foundation scour intelligent warning method according to claim 1, characterized in that: The warning threshold is specifically Δu=H / 75, wherein Δu is the horizontal displacement of the high-rise structure under the wind load, and H is the height of the high-rise structure.

7. An offshore wind turbine foundation scour intelligent early warning device, characterized in that: The computer program / instruction is executed by the processor to realize the steps of the offshore wind turbine foundation scour intelligent warning method in any one of claims 1-6.

8. A computer readable storage medium having stored thereon computer programs / instructions, characterized in that: The computer program / instruction is executed by the processor to realize the steps of the offshore wind turbine foundation scour intelligent warning method in any one of claims 1-6.

Images