Collaborative optimization design method and system for offshore wind turbine generator base

Abstract

The invention relates to the technical field of offshore wind power engineering, in particular to a collaborative optimization design method and system for an offshore wind turbine generator base, and the method comprises the steps: building an environment load model based on target sea area data, and generating a load working condition; a parameterized fan-base coupling model is constructed; aiming at minimizing the leveling cost per kilowatt-hour, and taking the structural strength and the inherent frequency as constraints; fan and base design variables are synchronously adjusted through multidisciplinary collaborative optimization, and candidate schemes are output; establishing a local high-precision model for the selected scheme to perform fatigue life verification, and if not, performing feedback optimization until a final scheme is obtained; and finally, based on the scheme, deploying the sensor and initializing the digital twin, realizing life prediction and graded early warning, and forming full life cycle closed-loop management.

Description

TECHNICAL FIELD

[0001] The present application relates to the offshore wind power engineering technical field, and particularly relates to a collaborative optimization design method and system of an offshore wind turbine generator set base. BACKGROUND

[0002] The offshore wind turbine generator set base is a key load-bearing structure supporting the entire wind turbine generator set, and its design needs to comprehensively consider complex environmental loads such as wind, wave, and current. At present, the industry generally adopts a sequential design method, that is, the wind turbine model and parameters are first determined according to experience, and then the supporting base structure is designed to match the wind turbine. In the design process, the structural engineer usually checks the ultimate strength and fatigue life of the base based on a simplified load model and standard working conditions to ensure its safety.

[0003] However, this traditional serial design mode has significant technical defects. The wind turbine and the base are designed as two relatively independent components, and there is a lack of means for global performance and cost collaborative optimization at the system level, which leads to the fact that the design scheme is often too conservative, the material redundancy is high, the foundation construction cost is high, or the dynamic coupling effect between the wind turbine and the foundation is not fully considered, which makes the structure have poor dynamic response under certain working conditions, and even has potential fatigue safety risks, ultimately affecting the overall economy and safety and reliability of the offshore wind power project. SUMMARY

[0004] In order to make up for the above shortcomings, the present application provides a collaborative optimization design method and system of an offshore wind turbine generator set base, which aims to improve the problem in the prior art that the scheme is often too conservative, the material redundancy is high, the foundation construction cost is high, or the dynamic coupling effect between the wind turbine and the foundation is not fully considered, which makes the structure have poor dynamic response under certain working conditions, and even has potential fatigue safety risks, ultimately affecting the overall economy and safety and reliability of the offshore wind power project.

[0005] In a first aspect, the present application provides the following technical scheme, a collaborative optimization design method of an offshore wind turbine generator set base, comprising the following steps:

[0006] Based on the marine hydrology and geology survey data of the target sea area, an environmental load model containing wind, wave, and current loads is established, and the environmental load model is used to generate a load working condition set containing extreme working conditions and fatigue working conditions;

[0007] Based on the preset wind turbine design parameters and base type, a parameterized wind turbine-base coupling model is constructed, and the parameters of the wind turbine-base coupling model include design variables related to the wind turbine design parameters and the base structure size;

[0008] The optimization objective is to minimize the levelized cost of electricity (LCOE) of the wind farm, and the constraints are the structural ultimate strength and the natural frequency avoiding the wind turbine excitation frequency range.

[0009] The load case set, the parameterized wind turbine-base coupling model, the optimization objective and the constraints are input into the optimization algorithm to perform multi-disciplinary collaborative optimization. The design variables related to the wind turbine design parameters and the base structure size are adjusted synchronously during the solution process to output one or more candidate design schemes.

[0010] Select a target design scheme from the candidate design schemes, extract key part information of the base structure based on the target design scheme, and establish a local high-precision finite element model including weld details for the key parts.

[0011] The fatigue load is applied to the local high-precision finite element model to calculate the fatigue life. If the fatigue life calculation result does not meet the design life requirements, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained.

[0012] Based on the final design scheme, strain sensors are installed at the key parts of the base structure, and a fatigue digital twin is initialized based on the final design scheme and the local high-precision finite element model to predict the remaining life and issue graded maintenance warnings.

[0013] Preferably, the process for establishing the environmental load model includes:

[0014] Acquire long-term historical observation data and numerical simulation data of the target sea area, including wind speed, wave height, wave period, current velocity and water depth data;

[0015] Based on the aforementioned long-term historical observation data and numerical simulation data, extreme environmental parameters with specific return periods are calculated using extreme value analysis methods.

[0016] Based on the aforementioned long-term historical observation data and numerical simulation data, a joint probability distribution model among wind, wave, and ocean current environmental parameters is established.

[0017] Based on the joint probability distribution model, multiple typical combinations of environmental parameters are generated through statistical sampling methods.

[0018] Based on the extreme environmental parameters, a set of extreme working condition loads for ultimate strength analysis is generated, and based on the typical combination of environmental parameters, a set of fatigue working condition loads for fatigue life analysis is generated.

[0019] Preferably, the construction process of the parameterized wind turbine-base coupling model includes:

[0020] Determine the wind turbine design parameters, which include rated power, rotor diameter, hub height, nacelle mass, and tower mass;

[0021] Determine the design variables related to the structural dimensions of the base based on the selected base type;

[0022] Based on the design variables related to the wind turbine design parameters and the base structure dimensions, a wind turbine-base coupled model is integrated and constructed on the simulation platform.

[0023] The design variables related to the wind turbine design parameters and the base structure dimensions are set as input parameters in the coupled model, so that the geometry and properties of the coupled model can be automatically updated as the input parameters change.

[0024] Preferably, the process, which sets the optimization objective as minimizing the levelized cost of electricity (LCOE) of the wind farm and uses structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints, includes:

[0025] An optimization objective function is established with the goal of minimizing the levelized cost of electricity (LCOE) of a wind farm. The LCOE is calculated based on the total investment cost of the wind farm, annual operation and maintenance costs, annual power generation, and discount rate.

[0026] The design variables related to the wind turbine design parameters and the base structure dimensions are mapped to the changes in the total investment cost of the wind farm.

[0027] The wind turbine design parameters are mapped to the changes in annual power generation;

[0028] Under extreme conditions set in the load case set, the structural response of the wind turbine-base coupling model must meet the preset strength safety criteria;

[0029] The first natural frequency of the wind turbine-base coupling model should be set to avoid the excitation frequency range formed by the rotation frequency of the wind turbine rotor and the passing frequency of the blades.

[0030] Preferably, the output process of the candidate design scheme includes:

[0031] Select an optimization algorithm to assign initial values ​​and change boundaries to the design variables related to the wind turbine design parameters and the base structure dimensions, and preset an optimization convergence condition;

[0032] Under the control of the optimization algorithm, the design variables related to the current value of the wind turbine design parameters and the base structure size are assigned to the parameterized wind turbine-base coupling model, and the load case set is called to perform calculations to obtain the performance index of the wind turbine-base coupling model under the optimization objective and the constraint conditions.

[0033] Based on the optimization algorithm, a judgment is made according to the performance index and the optimization convergence condition. If the optimization convergence condition is not met, a new set of design variable values ​​related to the wind turbine design parameters and the base structure size is generated synchronously. The new design variable values ​​related to the wind turbine design parameters and the base structure size are assigned to the parameterized wind turbine-base coupling model. The load case set is called again for calculation to obtain new performance indexes. The coupled system analysis and this optimization iteration decision-making steps are repeated.

[0034] When the optimization convergence condition is met, the design scheme corresponding to the design variables related to the current value of the wind turbine design parameters and the base structure dimensions is output as the candidate design scheme.

[0035] Preferably, the process for establishing the local high-precision finite element model includes:

[0036] Based on the target design scheme, through overall structural stress analysis, information on key parts of the base structure with significant stress concentration or dynamic response is extracted. The key parts include the connection area between the pile foundation and the transition section, and the pipe node weld area of ​​the jacket structure.

[0037] Based on the information of the key parts, a local three-dimensional solid geometric model including the geometry of the weld is established;

[0038] Assign material properties to the local three-dimensional solid geometry model, and set the displacement response of the key parts obtained from the wind turbine-base coupling model as the boundary condition of the local three-dimensional solid geometry model;

[0039] The local three-dimensional solid geometric model, which is given material properties and boundary conditions, is divided into high-density finite element meshes to generate the local high-precision finite element model.

[0040] Preferably, the process for obtaining the displacement response of the key components from the wind turbine-base coupling model includes:

[0041] In the parameterized wind turbine-base coupling model, one or more nodes corresponding to the key component information are located as boundary nodes.

[0042] The load case set is applied to the parameterized wind turbine-base coupling model, and the displacement response data of the boundary nodes under the load is calculated and obtained.

[0043] The displacement response data of the boundary nodes are used as a forced displacement constraint and applied to the corresponding region of the local three-dimensional solid geometry model.

[0044] Preferably, the fatigue life calculation process includes:

[0045] From the set of load conditions, extract the fatigue load spectrum for fatigue life analysis;

[0046] The fatigue load spectrum is applied to the local high-precision finite element model to calculate the structural stress time history of the key parts under the action of the fatigue load spectrum;

[0047] Based on the stress time history of the structure, the stress cycle is statistically analyzed using the rainflow counting method, and the cumulative fatigue damage of the key parts is calculated by combining the preset SN curve.

[0048] Based on the accumulated fatigue damage, the fatigue life of the critical components is calculated and evaluated.

[0049] Preferably, the operation process of the fatigue digital twin includes:

[0050] Based on the final design scheme, strain sensors are physically deployed at the key locations of the base structure.

[0051] On the computing platform, the geometry and properties of the digital twin are defined according to the final design scheme, and the fatigue digital twin is initialized using the local high-precision finite element model as its fatigue analysis engine.

[0052] Establish a data transmission link between the strain sensor and the fatigue digital twin;

[0053] Multiple fatigue damage thresholds are preset in the fatigue digital twin, and a corresponding maintenance warning level is associated with each fatigue damage threshold.

[0054] The fatigue digital twin continuously receives monitoring data from the strain sensor and calculates the cumulative fatigue damage of the key parts in real time based on the monitoring data and the local high-precision finite element model.

[0055] Based on the prediction of remaining lifespan based on the cumulative fatigue damage, when the cumulative fatigue damage reaches any of the fatigue damage thresholds, the corresponding maintenance warning level is triggered and issued.

[0056] Secondly, the present invention provides the following technical solution: a collaborative optimization design system for an offshore wind turbine base, the system comprising the following modules:

[0057] The environmental load modeling module is used to establish an environmental load model that includes wind, wave, and ocean current loads based on marine hydrological and geological survey data of the target sea area. The environmental load model is used to generate a load case set that includes extreme working conditions and fatigue working conditions.

[0058] The parametric model building module is used to build a parametric wind turbine-base coupling model based on preset wind turbine design parameters and base type. The parameters of the wind turbine-base coupling model include design variables related to the wind turbine design parameters and base structural dimensions.

[0059] The optimization problem definition module is used to set the optimization objective as minimizing the levelized cost of electricity (LCOE) of the wind farm, and to use the structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints.

[0060] The collaborative optimization solution module is used to input the load case set, the parameterized wind turbine-base coupling model, the optimization objective and the constraint conditions into the optimization algorithm to perform multi-disciplinary collaborative optimization solution. Through the solution process, the design variables related to the wind turbine design parameters and the base structure dimensions are adjusted synchronously, and one or more candidate design schemes are output.

[0061] The local fine-grained analysis module is used to select a target design scheme from the candidate design schemes, extract key part information of the base structure based on the target design scheme, and establish a local high-precision finite element model including weld details for the key parts.

[0062] The fatigue life verification and iteration module is used to apply fatigue load to the local high-precision finite element model and calculate fatigue life. If the fatigue life calculation result does not meet the design life requirements, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained.

[0063] The digital twin initialization and maintenance module is used to deploy strain sensors at the key parts of the base structure based on the final design scheme, and to initialize a fatigue digital twin based on the final design scheme and the local high-precision finite element model, in order to predict the remaining life and issue graded maintenance warnings.

[0064] The present invention has the following beneficial effects:

[0065] 1. In this invention, a parameterized wind turbine-base coupling model is constructed, and the core optimization objective is to minimize the levelized cost of electricity (LCOE) of the wind farm. The design parameters of the wind turbine and the structural dimensions of the base are used as design variables to be adjusted synchronously for multidisciplinary collaborative optimization. This allows for a systematic balance between power generation revenue and foundation construction costs while ensuring safety, thereby significantly reducing the overall cost and energy cost of the project and avoiding material waste caused by conservative design or potential power generation loss due to improper design.

[0066] 2. In this invention, by considering fatigue sensitivity during the collaborative optimization stage, a high-precision local finite element model containing weld details is established at the key stress concentration points for the optimized scheme to accurately verify fatigue life, which greatly improves the reliability of fatigue analysis. The verified accurate model is initialized as a digital twin, and fatigue damage is calculated and life is predicted online based on real-time monitoring data during the operation and maintenance stage. This enables proactive defense and predictive maintenance of hidden structural damage, greatly reducing the risk of structural failure due to fatigue fracture.

[0067] 3. In this invention, the parameterized fan-base coupling model enables the automatic adjustment of design variables and model updates, while the optimization algorithm replaces a large amount of tedious trial calculations by engineers, quickly selecting the optimal candidate scheme from a massive number of design schemes, greatly improving design efficiency. At the same time, by introducing a local high-precision finite element model for key parts, the fatigue life, a controlling factor, is thoroughly and specifically verified, ensuring that the final design scheme is not only optimal at the macroscopic system level, but also safe and reliable at the microscopic local stress level, effectively guaranteeing that the final output scheme has both high efficiency and high quality. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating a collaborative optimization design method for an offshore wind turbine base proposed in this invention.

[0069] Figure 2 This is a schematic diagram of the architecture of a collaborative optimization design system for an offshore wind turbine base proposed in this invention. Detailed Implementation

[0070] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] Example 1:

[0072] In a first embodiment of the present invention, the present invention provides a collaborative optimization design method for offshore wind turbine bases, such as... Figure 1 As shown, it includes the following steps:

[0073] Based on marine hydrological and geological survey data of the target sea area, an environmental load model including wind, wave, and ocean current loads is established. The environmental load model is used to generate a load case set including extreme and fatigue conditions.

[0074] Furthermore, the process for establishing the environmental load model includes:

[0075] Acquire long-term historical observation data and numerical simulation data of the target sea area, including wind speed, wave height, wave period, current velocity and water depth data;

[0076] Based on long-term historical observation data and numerical simulation data, extreme environmental parameters with specific return periods are calculated using extreme value analysis methods.

[0077] Based on long-term historical observation data and numerical simulation data, a joint probability distribution model of wind, wave, and ocean current environmental parameters is established.

[0078] Based on the joint probability distribution model, multiple typical combinations of environmental parameters are generated through statistical sampling methods.

[0079] Based on extreme environmental parameters, a set of extreme load conditions for ultimate strength analysis is generated, and based on typical combinations of environmental parameters, a set of fatigue load conditions for fatigue life analysis is generated.

[0080] Specifically, firstly, at least 20 years of long-term historical environmental data (adjustable by default) for the target sea area is collected. These data sources include field observation data and numerical simulation data. Field observation data comes from ocean buoys, wind towers, and ship reports, directly acquiring measured time series of wind speed, wave height, wave period, current velocity, current direction, and water depth. Numerical simulation data uses third-generation wave models (such as the SWAN model) and atmospheric models (such as the WRF model) for long-term numerical simulations, interpolating and completing the spatiotemporal gaps in the observation data to form a complete and continuous environmental database. Quality control is performed on all collected data, including removing obvious outliers, correcting systematic errors, and standardizing the time step and coordinate system. Extreme value analysis is used to calculate extreme environmental parameters for ultimate intensity verification. Specifically, a Poisson-Günber composite extreme value distribution model is used. Taking extreme wind speed as an example, the wind speed value of the largest independent storm event each year is selected from the long-term wind speed data to form an annual extreme value sample sequence. This sequence is fitted to a Günber distribution to calculate the extreme wind speeds encountered in the preset years. Similarly, using this as a design benchmark, the extreme effective wave height encountered within the preset lifespan can be calculated. and extreme flow rates These parameters together constitute the parameters of extreme environments;

[0081] To conduct fatigue analysis, a joint probability model of environmental factors such as wind, waves, and currents needs to be established. Here, a conditional modeling method is adopted, which uses significant wave height as a starting point. As the principal parameter, its long-term Weibull distribution is fitted, for each The intervals are then fitted with the corresponding wave cycles. The log-normal distribution of wind speed U and current velocity V, and the conditional probability distributions of current velocity V, are used to construct a complete (log-normal distribution). , A four-dimensional joint probability distribution model (U, V) is used to describe the probability of occurrence of all environmental states. Based on the above joint probability distribution model, more than 10,000 probabilistically representative "sea state blocks" are generated using Latin hypercube sampling techniques. These blocks represent typical combinations of environmental parameters, with each block containing one ( , (U, V) numerical pairs and their probabilities of occurrence;

[0082] The extreme environmental parameters obtained above ( , , Combined with different wind, wave, and current directions, 8-12 of the most dangerous load conditions are formed for subsequent ultimate strength analysis. Each of the thousands of sea state blocks generated above is transformed into a fatigue load condition. In addition to environmental parameters, each condition also includes its occurrence probability, which is used to calculate weighted damage in subsequent fatigue analysis. All fatigue conditions together constitute a complete long-term load spectrum.

[0083] Based on the preset wind turbine design parameters and base type, a parameterized wind turbine-base coupling model is constructed. The parameters of the wind turbine-base coupling model include design variables related to the wind turbine design parameters and base structural dimensions.

[0084] Furthermore, the construction process of the parameterized wind turbine-base coupled model includes:

[0085] Determine the wind turbine design parameters, which include rated power, rotor diameter, hub height, nacelle mass, and tower mass;

[0086] Determine the design variables related to the structural dimensions of the base based on the base type;

[0087] Based on the design variables related to the wind turbine design parameters and the base structure dimensions, a wind turbine-base coupled model is integrated and constructed on the simulation platform;

[0088] By setting the design variables related to the wind turbine design parameters and the base structure dimensions as input parameters in the coupled model, the geometry and properties of the coupled model can be automatically updated as the input parameters change.

[0089] Specifically, a set of core wind turbine design parameters are determined as optimization variables. These parameters include rated power, rotor diameter, hub height, and nacelle and tower mass. The base type, such as monopile foundation or jacket foundation, is selected based on project requirements and site conditions (e.g., water depth, geology). The corresponding structural dimension design variables are then determined. For monopile foundations, the main design variables include pile diameter (outer diameter of the steel pile), wall thickness (thickness of the steel pile wall, which can be defined in segments along the pile body), and penetration depth (the length of the pile driven below the seabed, the lower limit of which is determined by geological conditions). For jacket foundations, the main design variables include leg diameter and wall thickness, strut diameter and wall thickness, and total structural height and bottom span.

[0090] A coupled model can be constructed using the offshore wind power simulation software OpenFAST as the core simulation platform. OpenFAST has built-in capabilities for coupled aerodynamic-hydraulic-servo-elastic analysis. By customizing the elastic body module—specifically, using CAE tools (such as ANSYS or SACS) to parametrically generate a finite element model of the base structure—and integrating it with pre-defined models of the wind turbine tower, nacelle, and blades in OpenFAST, this integrated model can accurately simulate the aerodynamic loads of the wind turbine transmitted from the tower to the base structure, while simultaneously considering the complete load transfer path of wave and current dynamics on the base. This facilitates automation. Optimization requires fully parameterizing the model. In this embodiment, a Python script is written as the driving core. The script defines all the wind turbine and base parameters from the above steps as input variables. Based on the input variable values, the script can automatically call the API of the CAE software to regenerate the geometric and finite element model of the base structure, automatically update the relevant parameters (such as mass, stiffness, and hydrodynamic coefficients) in the OpenFAST input file, and drive OpenFAST to complete a full simulation calculation. In this way, the geometric, mass, stiffness, and damping properties of the coupled model can be automatically and dynamically updated as the input design variables change, eliminating the need for manual remodeling and improving automation.

[0091] The optimization objective is to minimize the levelized cost of electricity (LCOE) of the wind farm, and the constraints are the structural ultimate strength and the natural frequency avoiding the wind turbine excitation frequency range.

[0092] Furthermore, the process, which sets the optimization objective as minimizing the levelized cost of electricity (LCOE) of the wind farm and uses structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints, includes:

[0093] An optimization objective function is established with the goal of minimizing the levelized cost of electricity (LCOE) of wind farms. The LCOE is calculated based on the total investment cost of wind farms, annual operation and maintenance costs, annual power generation, and discount rate.

[0094] The design variables related to the wind turbine design parameters and the base structure dimensions are mapped to changes in the total investment cost of the wind farm.

[0095] Map the wind turbine design parameters to changes in annual power generation;

[0096] Under extreme conditions set in the load case set, the structural response of the wind turbine-base coupled model must meet the preset strength safety criteria;

[0097] The first natural frequency of the wind turbine-base coupling model should be set to avoid the excitation frequency range formed by the rotation frequency of the wind turbine rotor and the passing frequency of the blades.

[0098] Specifically, a mathematical model for calculating the levelized cost of electricity (LCOE) of a wind farm is established. The LCOE calculation formula adopts the industry-standard form, which is expressed as LCOE = (Total Investment Cost × CRF + Annual Operation and Maintenance Cost) / Annual Power Generation. The total investment cost includes the cost of wind turbine procurement, base material and manufacturing costs, transportation and installation costs, grid connection fees, etc. In this embodiment, a parameterized cost model is established to link costs with design variables, such as base cost. Can be modeled as ,in , and m are the fitting coefficients based on market data, and CRF is the capital recovery coefficient. The formula for calculating CRF is:

[0099] ;

[0100] in, For the discount rate, For the project lifecycle;

[0101] The annual operation and maintenance cost is estimated as a percentage of the total investment cost. The annual power generation is obtained by weighted summation of the power generation output of the wind turbine-base coupling model under fatigue load set. The power generation calculation takes into account the power curve of the wind turbine, wake loss and site wind resource distribution.

[0102] Using wind turbine design parameters (such as rated power and rotor diameter) and design variables related to the base structure dimensions (such as pile diameter and wall thickness) as inputs, the corresponding total investment cost change is calculated through the above-mentioned parametric cost model. Using wind turbine design parameters as inputs, the theoretical annual power generation is calculated by querying the wind turbine power curve and combining it with the site wind frequency distribution.

[0103] Under extreme load conditions within the load case set, the structural response of the wind turbine-base coupled model must meet a preset strength safety criterion. Specifically, in each iteration of the optimization algorithm, all extreme load conditions are applied to the coupled model, and the maximum equivalent stress of the entire structure is extracted. and the maximum deformation of key nodes Set the constraints as follows and ,in The yield strength of steel. For safety reasons, This represents the maximum allowable deformation according to the specifications.

[0104] The first natural frequency of the wind turbine-base coupling model must be set to avoid the wind turbine's excitation frequency range to prevent resonance. Specifically, based on the performance curve of the selected wind turbine model, the operating speed range of its rotor in the normal power generation area is determined. This range is usually provided by the wind turbine supplier. For example, for a typical modern variable speed wind turbine, its operating speed range may be 5 rpm to 12 rpm (revolutions per minute), i.e., the minimum operating speed. Maximum operating speed ;

[0105] The rotational frequency (1P frequency) range of the wind turbine rotor is obtained by converting the operating speed range, and the unit is Hertz (Hz), which can be expressed as:

[0106] ;

[0107] ;

[0108] Therefore, the frequency range of 1P is [0.083 Hz, 0.2 Hz][0.083Hz, 0.2Hz];

[0109] Blade passing frequency (3P frequency range): Since wind turbines typically have three blades, each blade passing over the tower causes a tower shadow effect and load disturbance when the rotor rotates, which can be expressed as:

[0110] ;

[0111] ;

[0112] Therefore, the 3P frequency range is [0.25 Hz, 0.6 Hz][0.25Hz, 0.6Hz];

[0113] To avoid resonance, the first natural frequency of the wind turbine-base coupling model is... The 1P and 3P frequency ranges mentioned above must be completely avoided, meaning the structure's natural frequency must be lower than the minimum excitation frequency. To ensure a safety margin, the constraints are set as follows:

[0114] ;

[0115] That is, the requirements are: .

[0116] The load case set, parameterized wind turbine-base coupling model, optimization objective and constraints are input into the optimization algorithm for multidisciplinary collaborative optimization. The design variables related to the wind turbine design parameters and base structure dimensions are adjusted synchronously during the solution process, and one or more candidate design schemes are output.

[0117] Furthermore, the output process for candidate design schemes includes:

[0118] Select an optimization algorithm, assign initial values ​​and change boundaries to the design variables related to the wind turbine design parameters and the base structure dimensions, and preset an optimization convergence condition;

[0119] Under the control of the optimization algorithm, the design variables related to the current value of the wind turbine design parameters and the base structure size are assigned to the parameterized wind turbine-base coupling model, and the load case set is called to perform calculations to obtain the performance index of the wind turbine-base coupling model under the optimization objective and constraint conditions.

[0120] Based on the optimization algorithm, the system judges the performance indicators and optimization convergence conditions. If the optimization convergence conditions are not met, a new set of design variable values ​​related to the wind turbine design parameters and base structure dimensions are generated simultaneously. The new design variable values ​​related to the wind turbine design parameters and base structure dimensions are assigned to the parameterized wind turbine-base coupling model. The load case set is called again for calculation to obtain new performance indicators. The coupled system analysis and this optimization iteration decision-making steps are repeated.

[0121] When the optimization convergence condition is met, the design schemes corresponding to the design variables related to the current value of the wind turbine design parameters and the base structure dimensions are output as candidate design schemes.

[0122] Specifically, this embodiment preferably uses a multi-objective genetic algorithm as the core optimizer, specifically the NSGA-II algorithm, which excels at handling constrained multi-objective optimization problems and can output a set of Pareto optimal solutions (i.e., a set of candidate design schemes). Initial values ​​are assigned to all wind turbine design parameters and design variables related to the base structure dimensions. For example, the initial value of the wind turbine's rated power is set to 10MW, and the initial value of the single pile diameter is set to 7 meters. These initial values ​​are set based on engineering experience, and reasonable upper and lower limits are set for each design variable to form an optimization search space. For example, the boundary of the single pile diameter can be set to [6m, 9m], and the boundary of the wall thickness can be set to [60mm, 90mm]. Two convergence criteria are set, and iteration stops when one is satisfied. One is to set the maximum number of iterations, for example, 200 generations, to prevent unlimited consumption of computational resources and ensure stability. The other is to ensure the stability of the solution, such as that the average improvement of the Pareto optimal solution set's front edge is less than 0.1% after 50 consecutive generations.

[0123] In each generation of the optimization algorithm, the following operations are performed on each individual in the population (i.e., a specific set of design variable values). This process is fully automated and driven by a Python main control script. The script automatically updates the geometry and properties of the wind turbine-base coupling model using the current individual's design variable values ​​through the aforementioned parameterized interface. It automatically calls the simulation software to apply each load case from the extreme load set to the updated model, calculates and records the maximum equivalent stress and maximum deformation of the entire structure, and automatically calls the simulation software to apply representative load cases from the fatigue load set to the updated model sequentially. It estimates the annual power generation (AEP) through weighted summation and calculates the first-order natural frequency of the updated model. It substitutes the current total investment cost, operation and maintenance cost, and annual power generation into the LCOE formula to calculate the specific value of the levelized cost of electricity (LCOE). It also calculates the violation of all constraints. For example, the strength constraint violation rate is set to max(0, (maximum equivalent stress / allowable stress) - 1), and the frequency constraint violation rate is set to max(0, (first-order frequency / 0.95 × 1). )-1);

[0124] The NSGA-II algorithm evaluates the fitness of each individual by performing non-dominated sorting and crowding calculation based on the objective function value (LCOE) and constraint violation degree of all individuals. The algorithm checks whether the preset convergence condition is met. If it does not converge, the algorithm generates a new set of design variable values ​​related to the wind turbine design parameters and base structure dimensions through genetic operations such as selection, crossover, and mutation, forming a new generation population. The new generation population returns to the coupled system analysis and optimization iteration decision-making steps, repeating the automated simulation analysis and performance index extraction process. When the optimization algorithm meets the convergence condition, the iteration stops. At this time, the Pareto optimal solution set output by the algorithm is a set of candidate design schemes. Each scheme contains a specific set of wind turbine and base design variable values, as well as corresponding performance indicators such as LCOE, strength, and frequency, for the decision-maker to make the final selection according to preference.

[0125] Select a target design scheme from the candidate design schemes. Based on the target design scheme, extract the key part information of the base structure. For the key parts, establish a local high-precision finite element model including weld details.

[0126] Furthermore, the process for establishing a local high-precision finite element model includes:

[0127] Based on the target design scheme, through overall structural stress analysis, information on key parts of the base structure with significant stress concentration or dynamic response is extracted. Key parts include the connection area between the pile foundation and the transition section, and the pipe node weld area of ​​the jacket structure.

[0128] Based on information about key components, a local three-dimensional solid geometric model containing the geometry of the weld is established;

[0129] Material properties are assigned to the local 3D solid geometry model, and the displacement response of key parts obtained from the wind turbine-base coupling model is set as the boundary condition of the local 3D solid geometry model.

[0130] A high-density finite element mesh is generated from the local three-dimensional solid geometric model that is given material properties and boundary conditions to produce a local high-precision finite element model.

[0131] Furthermore, the process of obtaining the displacement response of key components from the wind turbine-base coupling model includes:

[0132] In the parametric wind turbine-base coupling model, one or more nodes corresponding to the key component information are located as boundary nodes;

[0133] A set of load cases is applied to the parameterized wind turbine-base coupling model, and the displacement response data of the boundary nodes under load is calculated and obtained.

[0134] The displacement response data of the boundary nodes are used as forced displacement constraints and applied to the corresponding regions of the local 3D solid geometry model.

[0135] Specifically, from a set of candidate design schemes obtained through optimization, the design engineer selects a target design scheme based on a comprehensive consideration of factors such as LCOE and manufacturing process. This target design scheme includes a set of defined wind turbine design parameters and base structure dimensions. Using the aforementioned validated parameterized wind turbine-base coupling model corresponding to the target scheme, a detailed static and dynamic analysis is performed under extreme load conditions. By analyzing the stress cloud diagram, deformation diagram, and dynamic modal results of the overall model, areas with significant stress concentration or dynamic response are extracted. In this embodiment, for single pile foundations, the focus is on the connection weld area between the pile foundation and the transition section, which is a fatigue hotspot due to the abrupt change in cross-sectional stiffness. For jacket foundations, the focus is on K... For the weld area of ​​the T-shaped or Y-shaped pipe joint, a local geometry including key parts is cut out from the overall base model of the target scheme using professional finite element preprocessing software (such as ANSYS Workbench). The cutting range usually extends outward by at least 1.5 times the diameter of the main pipe to ensure that the boundary effect does not affect the stress results of the hot spot area. In the local geometric model, the geometry of the weld is established. In this embodiment, a solid modeling method is adopted. According to the relevant welding specifications (such as AWSD1.1), a weld entity with a specific weld leg size and cross-sectional shape (such as an isosceles triangle) is created and merged with the base material through Boolean operations to form a complete local three-dimensional solid geometric model containing weld details.

[0136] Assign material properties to the local 3D solid geometry model that are consistent with the overall model. For example, define the elastic modulus of steel as 210 GPa, Poisson's ratio as 0.3, and density as 7850 kg / m³. In the finite element mesh of the overall wind turbine-base coupled model, locate all nodes that cut the boundary around the critical parts and define these nodes as boundary nodes. Apply the most critical load cases (such as the design wave case) from the fatigue load set to the overall model, perform calculations, extract and save the displacement response data of all these boundary nodes under the load for six degrees of freedom (i.e., three translational displacements and three rotational displacements). Use the extracted boundary node displacement response data as the forced displacement... The displacement constraint is applied one-to-one to the corresponding nodes or surfaces of the local 3D solid geometry model, ensuring that the boundary conditions of the local model are completely consistent with those of the global model. The local 3D solid geometry model, which has been given material properties and boundary conditions, is meshed with a high-density finite element mesh using high-order solid elements (such as SOLID186). Significant mesh refinement is performed in the weld area and near the fusion line to ensure that the element size is small enough to capture steep stress gradients. For example, the element size at the weld toe is controlled at 2-3 mm. Finally, a local high-precision finite element model that can be used for high-precision stress analysis is generated. The number of meshes and the accuracy of this model are much higher than those of the global model.

[0137] The fatigue load is applied to a local high-precision finite element model to calculate the fatigue life. If the fatigue life calculation result does not meet the design life requirement, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained.

[0138] Furthermore, the fatigue life calculation process includes:

[0139] Extract the fatigue load spectrum for fatigue life analysis from the load case set.

[0140] The fatigue load spectrum is applied to a local high-precision finite element model to calculate the structural stress time history of key parts under the action of the fatigue load spectrum;

[0141] Based on the structural stress time history, the rainflow counting method is used to count stress cycles, and combined with the preset SN curve, the cumulative fatigue damage of key parts is calculated.

[0142] Based on cumulative fatigue damage, the fatigue life of critical components is calculated and evaluated.

[0143] Specifically, from the fatigue load set generated above, the load spectrum for analysis is extracted. This embodiment uses a spectral analysis method to balance computational efficiency and accuracy. It selects the top 50 worst sea states that contribute the most to the damage to key parts from thousands of sea state blocks. Each sea state block contains its environmental parameters ( , (U, V) and annual occurrence probability For each selected sea state, professional hydrodynamic and structural analysis software (such as ANSYS AQWA combined with Mechanical) is used to map the wave and current loads corresponding to the sea state onto the local high-precision finite element model. The model has been subjected to accurate boundary conditions through the sub-model method. Time-domain transient dynamic analysis is performed on each sea state, simulating at least 100 consecutive wave cycles to ensure statistical stability of the response. The analysis outputs the structural stress time history of the elements at the weld toe of the key part. It is particularly important to note here that the structural stress extracted is based on the equivalent nodal force, rather than the nominal stress, in order to eliminate mesh size sensitivity.

[0144] The structural stress time history data calculated for each sea state are imported into fatigue analysis software (such as nCodeDesignLife) or a corresponding algorithm library (such as Python's fatpack library). The stress time history is processed using the rainflow counting method to count all complete stress cycles and obtain the stress range for each stress cycle. and number of loops The SN curve for air-welded joints provided in the internationally recognized DNV GL-RP-C203 standard is adopted. For example, for pipe joints with welds, the D curve is selected, and its expression is: ,in = For each sea state i and each stress cycle j, calculate the fatigue damage caused by m=3. Considering the probability of this sea state occurring, the annual damage caused by this sea state is: The annual cumulative fatigue damage of key components is obtained by summing up the annual damage caused by all sea conditions. Then, the total cumulative fatigue damage within its design life T (e.g., 25 years) is... for: ;

[0145] Calculate the predicted fatigue life of this critical component. : To determine whether the fatigue life calculation results meet the design life requirements, according to the specifications, it is generally required that... ≤1 (i.e., the cumulative damage according to the Miner criterion is less than or equal to 1), if the requirement is met (i.e. If the target design scheme passes verification (≥25 years), it can be determined as the final design scheme (i.e., ...). If the fatigue life is less than 25 years, it indicates that the current design is insufficient in terms of fatigue performance. In this case, the design variables need to be readjusted. Specifically, return to the above-mentioned collaborative optimization solution steps, tighten the frequency constraints or add a penalty term for fatigue-sensitive design variables (such as wall thickness) in the optimization objective to guide the optimization algorithm to find a better solution. Then, re-execute the multidisciplinary collaborative optimization and all subsequent verification steps until a final design scheme that satisfies both the economic efficiency objective (LCOE) and all strength, frequency and fatigue life requirements is obtained.

[0146] Based on the final design scheme, strain sensors are installed in key parts of the base structure. Based on the final design scheme and local high-precision finite element model, a fatigue digital twin is initialized to predict the remaining life and issue graded maintenance warnings.

[0147] Furthermore, the operational process of the fatigue digital twin includes:

[0148] Based on the final design, strain sensors were physically deployed in key parts of the base structure.

[0149] On the computing platform, the geometry and properties of the digital twin are defined according to the final design scheme, and the fatigue digital twin is initialized using a local high-precision finite element model as its fatigue analysis engine.

[0150] Establish a data transmission link between the strain sensor and the fatigue digital twin;

[0151] Multiple fatigue damage thresholds are preset in the fatigue digital twin, and a corresponding maintenance warning level is associated with each fatigue damage threshold.

[0152] The fatigue digital twin continuously receives monitoring data from strain sensors and calculates the cumulative fatigue damage of key parts in real time based on the monitoring data and a local high-precision finite element model.

[0153] Based on the prediction of remaining life through cumulative fatigue damage, when the cumulative fatigue damage reaches any fatigue damage threshold, a corresponding maintenance warning level is triggered and issued.

[0154] Specifically, based on the final design drawings, sensors are deployed at key locations in the offshore wind turbine base structure (i.e., locations with stress concentration verified by fatigue life calculations, such as the mud surface of a monopile or the weld toe area of ​​the jacket pipe joint). Fiber optic strain sensors monitor the dynamic strain of the structure. Data acquisition instruments are installed inside the wind turbine tower or on the transition platform to power the sensors, demodulate the optical signals, and convert them into digital strain data. The acquisition instruments transmit the data in real time to a central server on shore or in the cloud via the wind farm's fiber optic network or wireless network. On the cloud server or high-performance computing platform, a fatigue digital twin is constructed and initialized. The 3D CAD model and material properties of the final design are directly imported as the geometric and physical basis of the digital twin. The verified local high-precision finite element model established during the above verification process and its complete set of calculation logic (including material constitutive model, boundary condition processing, and stress recovery algorithm) are encapsulated to form a callable computing service, which is the core of the digital twin—the fatigue analysis engine.

[0155] Develop a data interface program to establish a secure, low-latency data transmission link from the physical sensor data acquisition device to the cloud-based digital twin. The twin continuously monitors and receives real-time strain data streams from designated sensors. In the digital twin's management interface, multiple fatigue damage thresholds are preset, and each threshold is associated with a maintenance warning level. For example, a Level 1 warning (observation level) displays a yellow alert on the monitoring interface when the cumulative fatigue damage D reaches 0.3, indicating a need to pay attention to the degradation trend of that area. A Level 2 warning (planning level) issues an orange alert when D reaches 0.7 and automatically generates a report, suggesting... During the next planned downtime window, underwater robots will be deployed for close-range visual inspection or ultrasonic non-destructive testing. A Level 3 warning (action-level) will be issued. When D reaches 0.9, the system will issue a red alert, requiring the immediate development and execution of a detailed inspection and repair plan. After the digital twin initialization is complete, it will enter uninterrupted operation, executing the following automated process: The digital twin continuously receives real-time strain data, using it as load input to drive the integrated local high-precision finite element model for rapid stress calculation. Subsequently, using the same rainflow counting method and SN curve as in the design phase, real-time cumulative fatigue damage in key areas will be calculated online. Based on the current damage rate, predict the remaining fatigue life. , = The system will continue to Compare with preset thresholds at various levels, once If any threshold is reached or exceeded, the system will immediately and automatically trigger and send the corresponding maintenance warning level information to the upper-level management system and the mobile terminals of maintenance personnel. The information includes the wind turbine number, key parts, current damage level, predicted remaining life and recommended measures.

[0156] Example 2:

[0157] Traditional sequential design patterns suffer from significant technical drawbacks. The turbine and base are designed as two relatively independent components, lacking a means to perform global performance and cost optimization at the system level. This often leads to overly conservative designs with high material redundancy, increasing foundation construction costs. Furthermore, failure to fully consider the dynamic coupling effect between the turbine and base can result in poor dynamic response under certain operating conditions, and even potential fatigue safety risks, ultimately impacting the overall economic efficiency and reliability of offshore wind power projects. To address these issues, this invention provides a collaborative optimization design system for offshore wind turbine bases, the structure of which is as follows: Figure 2 As shown. The specific implementation process of this system is as follows:

[0158] The environmental load modeling module is used to establish an environmental load model that includes wind, wave, and ocean current loads based on marine hydrological and geological survey data of the target sea area. The environmental load model is used to generate a set of load conditions that include extreme and fatigue conditions.

[0159] The parametric model building module is used to build a parametric wind turbine-base coupling model based on preset wind turbine design parameters and base type. The parameters of the wind turbine-base coupling model include design variables related to wind turbine design parameters and base structural dimensions.

[0160] The optimization problem definition module is used to set the optimization objective as minimizing the levelized cost of electricity (LCOE) of the wind farm, and to use the structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints.

[0161] The collaborative optimization solution module is used to input the load case set, the parameterized wind turbine-base coupling model, the optimization objective and constraints into the optimization algorithm to perform multi-disciplinary collaborative optimization solution. Through the solution process, the design variables related to the wind turbine design parameters and the base structure dimensions are adjusted synchronously, and one or more candidate design schemes are output.

[0162] The local fine-grained analysis module is used to select a target design scheme from the candidate design schemes, extract key part information of the base structure based on the target design scheme, and establish a local high-precision finite element model including weld details for the key parts.

[0163] The fatigue life verification and iteration module is used to apply fatigue load to a local high-precision finite element model to calculate fatigue life. If the fatigue life calculation result does not meet the design life requirements, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained.

[0164] The digital twin initialization and maintenance module is used to deploy strain sensors in key parts of the base structure based on the final design scheme, and initialize a fatigue digital twin based on the final design scheme and local high-precision finite element model, which is used to predict the remaining life and issue graded maintenance warnings.

[0165] Specifically, the environmental load modeling module first constructs a load set including extreme and fatigue conditions based on site data, providing input for subsequent analysis. The parametric model construction module simultaneously creates a parametric coupling model integrating the wind turbine and the base, defining the optimized design space. On this basis, the optimization problem definition module sets the minimization of levelized cost of electricity (LCOE) as the economic objective, with strength and frequency as safety constraints. The co-optimization solution module then drives the optimization algorithm, simultaneously adjusting the design variables of the wind turbine and the base, automatically finding the best option and outputting candidate design schemes. Subsequently, the local refined analysis module conducts in-depth verification of the preferred scheme, establishes high-precision models of key parts, and the fatigue life verification and iteration module completes accurate fatigue life calculations. If the results do not meet the requirements, they are fed back to the optimization module for iteration until a safe, economical, and durable final design scheme is obtained. Finally, the digital twin initialization and operation and maintenance module materializes the design scheme and refined model, deploys sensors, and initializes the digital twin, realizing real-time fatigue damage monitoring, remaining life prediction, and graded early warning based on real data, thus forming a technical closed loop from virtual precise design to physical predictive operation and maintenance.

[0166] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A collaborative optimization design method for the base of an offshore wind turbine, characterized in that, Includes the following steps: Based on marine hydrological and geological survey data of the target sea area, an environmental load model including wind, wave, and ocean current loads is established. The environmental load model is used to generate a load condition set including extreme working conditions and fatigue working conditions. Based on preset wind turbine design parameters and base type, a parameterized wind turbine-base coupling model is constructed. The parameters of the wind turbine-base coupling model include design variables related to wind turbine design parameters and base structural dimensions. The optimization objective is to minimize the levelized cost of electricity (LCOE) of the wind farm, and the constraints are the structural ultimate strength and the natural frequency avoiding the wind turbine excitation frequency range. The load case set, the parameterized wind turbine-base coupling model, the optimization objective and the constraints are input into the optimization algorithm to perform multi-disciplinary collaborative optimization. The design variables related to the wind turbine design parameters and the base structure size are adjusted synchronously during the solution process to output one or more candidate design schemes. Select a target design scheme from the candidate design schemes, extract key part information of the base structure based on the target design scheme, and establish a local high-precision finite element model including weld details for the key parts. The fatigue load is applied to the local high-precision finite element model to calculate the fatigue life. If the fatigue life calculation result does not meet the design life requirements, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained. Based on the final design scheme, strain sensors are installed at the key parts of the base structure, and a fatigue digital twin is initialized based on the final design scheme and the local high-precision finite element model to predict the remaining life and issue graded maintenance warnings.

2. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The process of establishing the environmental load model includes: Acquire long-term historical observation data and numerical simulation data of the target sea area, including wind speed, wave height, wave period, current velocity and water depth data; Based on the aforementioned long-term historical observation data and numerical simulation data, extreme environmental parameters with specific return periods are calculated using extreme value analysis methods. Based on the aforementioned long-term historical observation data and numerical simulation data, a joint probability distribution model among wind, wave, and ocean current environmental parameters is established. Based on the joint probability distribution model, multiple typical combinations of environmental parameters are generated through statistical sampling methods. Based on the extreme environmental parameters, a set of extreme working condition loads for ultimate strength analysis is generated, and based on the typical combination of environmental parameters, a set of fatigue working condition loads for fatigue life analysis is generated.

3. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The construction process of the parameterized wind turbine-base coupling model includes: Determine the wind turbine design parameters, which include rated power, rotor diameter, hub height, nacelle mass, and tower mass; Determine the design variables related to the structural dimensions of the base based on the selected base type; Based on the design variables related to the wind turbine design parameters and the base structure dimensions, a wind turbine-base coupled model is integrated and constructed on the simulation platform. The design variables related to the wind turbine design parameters and the base structure dimensions are set as input parameters in the coupled model, so that the geometry and properties of the coupled model can be automatically updated as the input parameters change.

4. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The process, which sets the optimization objective as minimizing the levelized cost of electricity (LCOE) of a wind farm and uses structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints, includes: An optimization objective function is established with the goal of minimizing the levelized cost of electricity (LCOE) of a wind farm. The LCOE is calculated based on the total investment cost of the wind farm, annual operation and maintenance costs, annual power generation, and discount rate. The design variables related to the wind turbine design parameters and the base structure dimensions are mapped to the changes in the total investment cost of the wind farm. The wind turbine design parameters are mapped to the changes in annual power generation; Under extreme conditions set in the load case set, the structural response of the wind turbine-base coupling model must meet the preset strength safety criteria; The first natural frequency of the wind turbine-base coupling model should be set to avoid the excitation frequency range formed by the rotation frequency of the wind turbine rotor and the passing frequency of the blades.

5. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The output process for the candidate design scheme includes: Select an optimization algorithm to assign initial values ​​and change boundaries to the design variables related to the wind turbine design parameters and the base structure dimensions, and preset an optimization convergence condition; Under the control of the optimization algorithm, the design variables related to the current value of the wind turbine design parameters and the base structure size are assigned to the parameterized wind turbine-base coupling model, and the load case set is called to perform calculations to obtain the performance index of the wind turbine-base coupling model under the optimization objective and the constraint conditions. Based on the optimization algorithm, a judgment is made according to the performance index and the optimization convergence condition. If the optimization convergence condition is not met, a new set of design variable values ​​related to the wind turbine design parameters and the base structure size is generated synchronously. The new design variable values ​​related to the wind turbine design parameters and the base structure size are assigned to the parameterized wind turbine-base coupling model. The load case set is called again for calculation to obtain new performance indexes. The coupled system analysis and this optimization iteration decision-making steps are repeated. When the optimization convergence condition is met, the design scheme corresponding to the design variables related to the current value of the wind turbine design parameters and the base structure dimensions is output as the candidate design scheme.

6. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The process of establishing the local high-precision finite element model includes: Based on the target design scheme, through overall structural stress analysis, information on key parts of the base structure with significant stress concentration or dynamic response is extracted. The key parts include the connection area between the pile foundation and the transition section, and the pipe node weld area of ​​the jacket structure. Based on the information of the key parts, a local three-dimensional solid geometric model including the geometry of the weld is established; Assign material properties to the local three-dimensional solid geometry model, and set the displacement response of the key parts obtained from the wind turbine-base coupling model as the boundary condition of the local three-dimensional solid geometry model; The local three-dimensional solid geometric model, which is given material properties and boundary conditions, is meshed with a high-density finite element mesh to generate the local high-precision finite element model.

7. The collaborative optimization design method for an offshore wind turbine base according to claim 6, characterized in that, The process of obtaining the displacement response of the key components from the wind turbine-base coupling model includes: In the parameterized wind turbine-base coupling model, one or more nodes corresponding to the key component information are located as boundary nodes. The load case set is applied to the parameterized wind turbine-base coupling model, and the displacement response data of the boundary nodes under the load is calculated and obtained. The displacement response data of the boundary nodes are used as a forced displacement constraint and applied to the corresponding region of the local three-dimensional solid geometry model.

8. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The fatigue life calculation process includes: From the set of load conditions, extract the fatigue load spectrum for fatigue life analysis; The fatigue load spectrum is applied to the local high-precision finite element model to calculate the structural stress time history of the key parts under the action of the fatigue load spectrum; Based on the stress time history of the structure, the stress cycle is statistically analyzed using the rainflow counting method, and the cumulative fatigue damage of the key parts is calculated by combining the preset SN curve. Based on the accumulated fatigue damage, the fatigue life of the critical components is calculated and evaluated.

9. The collaborative optimization design method for an offshore wind turbine base according to claim 1, characterized in that, The operation process of the fatigue digital twin includes: Based on the final design scheme, strain sensors are physically deployed at the key locations of the base structure. On the computing platform, the geometry and properties of the digital twin are defined according to the final design scheme, and the fatigue digital twin is initialized using the local high-precision finite element model as its fatigue analysis engine. Establish a data transmission link between the strain sensor and the fatigue digital twin; Multiple fatigue damage thresholds are preset in the fatigue digital twin, and a corresponding maintenance warning level is associated with each fatigue damage threshold. The fatigue digital twin continuously receives monitoring data from the strain sensor and calculates the cumulative fatigue damage of the key parts in real time based on the monitoring data and the local high-precision finite element model. Based on the prediction of remaining lifespan based on the cumulative fatigue damage, when the cumulative fatigue damage reaches any of the fatigue damage thresholds, the corresponding maintenance warning level is triggered and issued.

10. A collaborative optimization design system for an offshore wind turbine base, characterized in that, The collaborative optimization design method for an offshore wind turbine base as described in any one of claims 1-9, the system comprising the following modules: The environmental load modeling module is used to establish an environmental load model that includes wind, wave, and ocean current loads based on marine hydrological and geological survey data of the target sea area. The environmental load model is used to generate a load case set that includes extreme working conditions and fatigue working conditions. The parametric model building module is used to build a parametric wind turbine-base coupling model based on preset wind turbine design parameters and base type. The parameters of the wind turbine-base coupling model include design variables related to the wind turbine design parameters and base structural dimensions. The optimization problem definition module is used to set the optimization objective as minimizing the levelized cost of electricity (LCOE) of the wind farm, and to use the structural ultimate strength and natural frequency avoiding the wind turbine excitation frequency range as constraints. The collaborative optimization solution module is used to input the load case set, the parameterized wind turbine-base coupling model, the optimization objective and the constraint conditions into the optimization algorithm to perform multi-disciplinary collaborative optimization solution. Through the solution process, the design variables related to the wind turbine design parameters and the base structure dimensions are adjusted synchronously, and one or more candidate design schemes are output. The local fine-grained analysis module is used to select a target design scheme from the candidate design schemes, extract key part information of the base structure based on the target design scheme, and establish a local high-precision finite element model including weld details for the key parts. The fatigue life verification and iteration module is used to apply fatigue load to the local high-precision finite element model and calculate fatigue life. If the fatigue life calculation result does not meet the design life requirements, the design variables are readjusted and the multidisciplinary collaborative optimization solution and subsequent steps are re-executed until all design requirements are met and the final design scheme is obtained. The digital twin initialization and maintenance module is used to deploy strain sensors at the key parts of the base structure based on the final design scheme, and to initialize a fatigue digital twin based on the final design scheme and the local high-precision finite element model, in order to predict the remaining life and issue graded maintenance warnings.

Images