Tension leg platform coupled motion response and hull parameter optimization method and system

Abstract

The application belongs to the technical field of ocean engineering, and provides a tension leg platform coupling motion response and main body parameter optimization method and system. The application establishes a model of wind wheel unsteady aerodynamic load, platform six-degree-of-freedom hydrodynamic force and mooring system nonlinear tension characteristics, introduces an agent model to establish the relationship between design variables and objective functions in this paper, takes platform geometric dimensions as design variables, finds the optimal design of platform configuration which meets the maximum average power generation, minimum power generation standard deviation and other targets, and realizes comprehensive optimization. The application reduces the variables such as column spacing and column radius which need to be determined through multiple trial calculations into one directional optimization, synchronously reduces platform power fluctuation and steel consumption, and provides a directly reusable design path for weight reduction and rapid iteration of the tension leg platform.

Description

Technical Field

[0001] This invention belongs to the field of marine engineering technology, specifically relating to a method and system for optimizing the coupled motion response and main parameters of a tension leg floating platform. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Existing tension leg platform designs mostly adopt fixed master dimensions and empirical wall thickness allocation, relying on repeated calculations to balance motion performance and steel consumption. It is difficult to achieve a quantitative balance between power generation fluctuations, structural weight and construction costs. On the other hand, existing optimization cycles are long and iteration costs are high. Summary of the Invention

[0004] To address the aforementioned problems, this invention proposes a method and system for optimizing the coupled motion response and main parameters of a tension leg floating platform. This invention models the unsteady aerodynamic loads of the wind turbine, the six-degree-of-freedom hydrodynamics of the platform, and the nonlinear tension characteristics of the mooring system. It also introduces a surrogate model to establish the relationship between the design variables and the objective function. Using the platform's geometric dimensions as design variables, the invention seeks the optimal platform configuration that satisfies objectives such as the maximum average power generation and the minimum standard deviation of power generation, thus achieving comprehensive optimization.

[0005] According to some embodiments, the present invention adopts the following technical solution: A method for optimizing the coupled motion response and main parameters of a tension leg floating platform includes the following steps: Based on the geometry and parameters of the tension leg floating platform, a model of the tension leg floating platform is constructed. Frequency domain calculations are performed based on the tension leg floating platform model; Based on the design scheme of the tension leg floating platform, the layout of the mooring system and the mooring line material are determined. A mooring line model is established based on the tension leg floating platform model, and a static analysis is performed on the mooring line model. Based on the static analysis results, the frequency domain calculation results were corrected. Based on the corrected frequency domain calculation results and the mooring line model, and combined with the wind turbine parameters, a wind turbine model is established. Based on the wind turbine model and mooring line model, aerodynamic-hydraulic-mooring coupling analysis is performed; Based on the coupling analysis results, the key geometric parameters of the tension leg floating platform were selected as optimization variables. A Kriging model was used to construct a proxy model of the objective function for each input variable. The optimization results were obtained by using a genetic algorithm.

[0006] As an alternative implementation method, the process of constructing a tension leg floating platform model based on the geometry and parameters of the tension leg floating platform includes: determining the dimensions of each column, pontoon, and brace, as well as the column spacing, based on the geometry and main dimensions of the tension leg floating platform; calculating the total mass, center of mass, moment of inertia, draft, and displacement of the platform; constructing a three-dimensional solid element model; and meshing the model.

[0007] As an alternative implementation method, the process of performing frequency domain calculations based on the tension leg floating platform model includes: using a frequency domain solver, calculating the platform's six-degree-of-freedom additional mass, radiation damping, hydrostatic restoring stiffness, and unit amplitude excitation force based on the tension leg floating platform model, and further obtaining the loads and displacements; The platform's dynamic response frequency domain equation of motion is:

[0008]

[0009] In the formula: For frequency, The Fourier transform of the platform's time-domain motion response; The system mass matrix includes the hydrodynamic added mass matrix caused by radiation. and structural mass matrix ; Here is the damping matrix; This is the stiffness matrix, including the mooring stiffness matrix at the static equilibrium position of the structure. and the hydrostatic stiffness matrix of the floating structure ; This is the environmental load matrix, including aerodynamic loads. First-order wave excitation force and second-order low-frequency wave force .

[0010] As an alternative implementation method, the process of establishing a mooring line model includes: Based on the platform design, the layout of the mooring system and the materials of the mooring line are determined. The tension leg floating platform model and frequency domain calculation results are imported into the time-domain hydrodynamic analysis software, and the mooring line model is established using the lumped mass method. In the lumped mass method, the performance of the mooring line is equivalent to a nonlinear spring, consisting of several continuous, massless segments and nodes at the midpoints of each segment. Each segment is a continuous, massless pipeline element, and its axial, bending, and torsional characteristics are considered only. It is simulated as a combination of axial, bending, and torsional spring dampers. The nodes concentrate half the mass of two adjacent segments, and all the distributed forces acting on the pipeline element are concentrated at its concentrated mass point. The effective tension on the pipeline is obtained by stress-strain relationship according to the transient position of the node.

[0011] As an alternative implementation, the process of correcting the frequency domain calculation results based on the static analysis results includes: performing static analysis on the platform and mooring system to obtain the platform's mooring stiffness matrix, which is used to characterize the platform's restoring force response under six degrees of freedom and small displacement; using the derived stiffness matrix as the external constraint of the frequency domain solver; repeating the frequency domain analysis steps; and correcting the frequency domain calculation results.

[0012] As an alternative implementation method, the process of establishing a wind turbine model includes establishing a wind turbine model based on parameters such as rotor radius, airfoil lift-drag coefficient, blade mass distribution, and tower height, according to the corrected calculation results and mooring line model. Among them, the wind turbine blades are modeled using blade element momentum theory, the tower is modeled using the lumped mass method, and the nacelle and hub are modeled as six-degree-of-freedom solids.

[0013] As an alternative implementation method, the process of performing aerodynamic-hydraulic-mooring coupling analysis includes: calculating the aerodynamic load of the wind turbine using blade element momentum theory, and the lift coefficient... and drag coefficient The expression is as follows:

[0014] In the formula and For the projections of lift and drag onto the z-axis and x-axis, Indicates the relative velocity of the incoming flow. The air density is given; the lift coefficient and drag coefficient are related to the airfoil, Reynolds number, and angle of attack; when the lift coefficient and drag coefficient are known, the lift and drag on each airfoil are solved by airfoil parameters, and the load is obtained by integrating over all airfoils of the entire blade; during the solution process, an external link library is used to control the blade pitch and rotor speed, the time-domain analysis software outputs the blade position information and aerodynamic loads, and the external link library updates the blade pitch and rotor speed; The platform's hydrodynamic response is calculated using a hybrid method combining three-dimensional potential flow theory and the Morison equation. For the main platform structure larger than a set size, diffraction / radiation theory is employed, with time-domain solutions derived from corrected frequency-domain hydrodynamic coefficients. For structures smaller than the set size, the Morison equation is used for calculation.

[0015] As an alternative implementation method, a Kriging model is used to construct a proxy model of the objective function for each input variable. The process includes selecting the column radius, column spacing, and buoy radius as optimization variables. The Kriging model consists of a multinomial regression term describing the overall trend of the response and a local bias term, and its calculation formula is as follows:

[0016] In the formula, , Regression models and normally distributed Gaussian stochastic processes; To optimize variables; These are the regression coefficients; The covariance expression is as follows:

[0017] In the formula, To indicate and The related functions are expressed as follows:

[0018] In the formula, Spatial dimension; For the relevant function value to follow The rate of change; For the smoothness of the model; The relevant undetermined parameters of the Kriging model were determined through maximum likelihood estimation, and the unknown points... The best linear unbiased estimate is:

[0019] In the formula, for Least squares estimate; The vector representing the relationship between unknown and known points; for The mean squared error; This is the estimated value of the target by the proxy model; Latin hypercube sampling is used to generate initial sample points in the three-dimensional design space to ensure that the samples are uniformly distributed throughout the domain and avoid clustering. Each sample point is then substituted into the aerodynamic-hydraulic-mooring coupling analysis of the floating wind turbine to calculate the platform's six-degree-of-freedom motion response, mooring tension extremes, and power generation time history. The mean and standard deviation of the power are extracted to form a sample library. Based on the base sample library, a surrogate model for each objective function on each input variable is established using the Kriging model.

[0020] As an alternative implementation method, the process of using a genetic algorithm for optimization includes: Using the established 3D design space as the search domain for decision variables, upper and lower bounds are set for each variable; the random generation scale is... The initial parent population Each individual represents a set of platform geometric scales; an algebraic counter is set. ; Will All individuals are fed into a trained and validated Kriging agent model; the output for each individual is a bi-objective value: the mean power generation. With power standard deviation To form the target matrix ; right Perform a non-dominated sort, which yields X non-dominated fronts. , ,..., This indicates that the parameters are obtained after performing a non-dominated sort on each individual in the population. , indicating the dominance of an individual The number of individuals in the population Individuals are denoted as non-dominated frontiers ,remember The individual's rank value is 1; Other individuals in The rules are assigned to the next level of non-dominated front. In each iteration, all individuals in the population are traversed, and a corresponding non-dominated front is assigned to each individual. The crowding degree is calculated among individuals with the same rank value. The individual with the smallest crowding degree coefficient in the entire solution set is deleted to maintain population diversity. A binary tournament selection method is adopted: individuals with low non-dominance level and high crowding are given priority to be retained in the next generation, and the real numbers are encoded using the simulated binary crossover SBX operator. Crossover is performed using the SBX operator, and polynomial mutation is used to generate N new individuals, which form the offspring population. ; Call the Kriging proxy model again to obtain Target value; By merging the parent and child generations, we obtain ;right Perform non-dominated ranking, selecting individuals with low non-dominated rank and high crowding to form the next generation of parents. Termination condition: If Output the non-dominated solution set and end the loop; otherwise, merge the parent and child generations, repeat the selection-crossover-mutation operation in S54, and generate a new child generation. And its target value was evaluated using a Kriging model; .

[0021] A tension leg floating platform coupled motion response and main body parameter optimization system includes: The tension leg floating platform model building module is configured to build a tension leg floating platform model based on the geometry and parameters of the tension leg floating platform. The frequency domain calculation module is configured to perform frequency domain calculations based on the tension leg floating platform model; The mooring line model construction module is configured to determine the layout of the mooring system and the mooring line material based on the tension leg floating platform design scheme, build a mooring line model based on the tension leg floating platform model, and perform static analysis on the mooring line model. The correction module is configured to correct the frequency domain calculation results based on the static analysis results; The wind turbine model building module is configured to build a wind turbine model based on the corrected frequency domain calculation results and the mooring line model, combined with the wind turbine parameters. The coupling analysis module is configured to perform aerodynamic-hydraulic-mooring coupling analysis based on the wind turbine model and the mooring line model; The optimization solution module is configured to select the key geometric parameters of the tension leg floating platform as optimization variables based on the coupling analysis results, construct a surrogate model of the objective function for each input variable using a Kriging model, and use a genetic algorithm to perform optimization and obtain the optimization results.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention realizes fully automatic multi-objective optimization of the main scale of the tension leg platform: by embedding the maximum mean power generation and the minimum power standard deviation into the objective function, the surrogate-evolutionary coupling algorithm can simultaneously optimize the platform geometry, ballast and mooring parameters, and iteratively output quantitative scale combinations without the need for repeated manual calculations, thus transforming the platform design from "experience-based trial and error" to "directed optimization".

[0023] This invention enhances the ability to understand the coupling relationship between complex environments and design variables: the state-target mapping is explicitly expressed by the Kriging model, which can reveal the sensitivity ranking of the platform's main scale to motion and load, providing an interpretable quantitative basis for subsequent detailed design.

[0024] This invention optimizes the design parameters of the tension leg platform using NSGA-II, reducing variables such as column spacing and column radius, which originally required multiple rounds of trial calculations, to a single directional optimization. This reduces platform power fluctuations and steel consumption in sync, providing a directly reusable design path for weight reduction and rapid iteration of the tension leg platform.

[0025] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0026] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0027] Figure 1 This is a schematic diagram of the anchor chain modeling based on the lumped mass method of the present invention, wherein (a) is a schematic diagram of the mooring line structure and (b) is a schematic diagram simulating a spring; Figure 2 This is a flowchart of the platform size parameter steps based on the NSGA-II algorithm of this invention; Figure 3 This is a schematic diagram of the tension leg platform for which this invention is based. Detailed Implementation

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0029] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0030] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0031] Where there is no conflict, the embodiments and features described in this application may be combined with each other.

[0032] Example 1 A method for calculating the motion response of a tension leg wind turbine and a platform optimization design method based on an aerodynamic-hydraulic-mooring coupling model, comprising the following steps: S1: Platform Main Parametric Modeling Create a 3D solid element model of the platform in a CAD / CAE environment, such as Figure 3 As shown, the model includes the platform's geometric configuration and main dimensions, determining the dimensions of each column, pontoon, and brace, as well as the column spacing; calculating the platform's total mass, center of mass position, moment of inertia, draft, and displacement; meshing the model to ensure that all shell elements form a single closed wetted surface; and outputting the model in a format that can be imported into hydrodynamic analysis software (such as .hst, .gdf) for subsequent use by frequency domain and time domain software.

[0033] S2: Frequency Domain Analysis Import the model established in S1 into a professional frequency domain solver (such as OrcaWave or WAMIT) to output the platform's six-degree-of-freedom additional mass, radiation damping, hydrostatic restoring stiffness, and unit amplitude excitation force, and further obtain the load RAO and displacement RAO.

[0034] The frequency domain equation of motion for the dynamic response of a floating platform can be written as:

[0035]

[0036] In the formula: For frequency, This is the Fourier transform of the platform's time-domain motion response. The system mass matrix includes the hydrodynamic added mass matrix caused by radiation. and structural mass matrix . Here is the damping matrix. This is the stiffness matrix, including the mooring stiffness matrix at the static equilibrium position of the structure. and the hydrostatic stiffness matrix of the floating structure . This is the environmental load matrix, including aerodynamic loads. First-order wave excitation force and second-order low-frequency wave force .

[0037] S3: Mooring System Design and Static Analysis Based on the platform design, determine the layout of the mooring system and the materials of the mooring line. Import the results of S2 into time-domain hydrodynamic analysis software (such as OrcaFlex) to build the platform model, and use the lumped mass method to build the mooring line model.

[0038] In the lumped quality method, such as Figure 1 As shown in (a), the mooring line's performance is equivalent to a nonlinear spring, consisting of several continuous, massless segments and nodes at the midpoints of each segment. Each segment is a continuous, massless pipeline element, considering only its axial, bending, and torsional characteristics, such as... Figure 1 As shown in (b), it is simulated as a combination of axial, bending, and torsional spring dampers. Each node concentrates half the mass of two adjacent segments, and all distributed forces acting on the pipeline unit (fluid damping force, inertial force, internal damping force, gravity, buoyancy, etc.) are considered to be concentrated at their concentrated mass point. The effective tension in the pipeline... The stress-strain relationship is obtained based on the transient position of the node.

[0039]

[0040] In the formula, It is wall tension. and For external and internal pressures, and It represents the stress area of ​​the external and internal cross-sections. A negative effective tension indicates compression. For linear axial stiffness, Defined as:

[0041]

[0042] In the formula, It is axial stiffness. It is the total average axial strain. It is the instantaneous length of the line segment. It is the line segment stretching coefficient. Unstretched line segment length Poisson's ratio, This is the tension-torque coupling value. The twist angle of the line segment.

[0043] Static analysis was performed on the platform and mooring system to obtain the platform's mooring stiffness matrix. The stiffness matrix is ​​used to characterize the restoring force (moment) response of the platform under small displacements of six degrees of freedom. As an external constraint for the frequency domain solver, repeat step S2 to correct the results of the frequency domain calculation.

[0044] S4: Overall Machine Modeling and Aerodynamic-Hydrodynamic-Mooring Coupling Analysis Import the S3-corrected calculation results and mooring line model into time-domain hydrodynamic analysis software (such as OrcaFlex). Establish a wind turbine model based on parameters such as rotor radius, airfoil lift and drag coefficient, blade mass distribution, and tower height. The wind turbine blades are modeled using BEM (blade element momentum theory), the tower is modeled using the lumped mass method, and the nacelle and hub are modeled as six-degree-of-freedom solids.

[0045] The aerodynamic load of the wind turbine is calculated using blade element momentum theory, and the lift coefficient is... and drag coefficient The expression is as follows:

[0046] In the formula and For the projections of lift and drag onto the z-axis and x-axis, Indicates the relative velocity of the incoming flow. The air density is given. The lift and drag coefficients are related to the airfoil, Reynolds number, and angle of attack. When the lift and drag coefficients are known, the lift and drag on each airfoil are solved using airfoil parameters. The loads can be obtained by integrating over all airfoils of the entire blade. During the solution process, an external library is used to control the blade pitch and rotor speed. The time-domain analysis software outputs blade position information and aerodynamic loads, while the external library updates the blade pitch and rotor speed.

[0047] The platform's hydrodynamic response is calculated using a hybrid method combining three-dimensional potential flow theory and the Morison equation. For large-scale platform structures (such as columns), diffraction / radiation theory is employed, with time-domain solutions derived from S3-corrected frequency-domain hydrodynamic coefficients. For small-scale structures (such as braces), the Morison equation is used, as shown in the following formula:

[0048] In the formula, It is the wave force acting on the structure; It is the instantaneous wave fluid velocity perpendicular to the object's axis; It is the instantaneous wave-fluid acceleration perpendicular to the axis of the object; It is the drag coefficient; It is the inertial force coefficient.

[0049] S5: Platform Geometry Parameter Optimization Geometric parameters of the platform, such as column radius, column spacing, and pontoon radius, are selected as optimization variables, and a surrogate model is constructed using a Kriging model.

[0050] The Kriging model consists of a multinomial regression term describing the overall trend of the response and a local bias term, and its calculation formula is as follows:

[0051] In the formula, , Regression models and normally distributed Gaussian stochastic processes; To optimize variables; is the regression coefficient. The covariance expression is as follows:

[0052] In the formula, To indicate and The related functions are expressed as follows:

[0053] In the formula, Spatial dimension; For the relevant function value to follow The rate of change; This refers to the smoothness of the model.

[0054] The relevant undetermined parameters of the Kriging model can be determined through maximum likelihood estimation. Unknown points. The best linear unbiased estimate is:

[0055] In the formula, for Least squares estimate; The vector representing the relationship between unknown and known points; for The mean squared error; This is the estimated value of the target by the proxy model.

[0056] Latin hypercube sampling is used to generate initial sample points in the three-dimensional design space to ensure that the samples are uniformly distributed throughout the domain and avoid clustering. The sample points are then substituted one by one into the above-mentioned floating wind turbine aerodynamic-hydraulic-mooring coupled numerical model to calculate the platform's six-degree-of-freedom motion response, mooring tension extremes, and power generation time history. The mean and standard deviation of the power are extracted to form a sample library.

[0057] Based on the above sample database, a surrogate model for each objective function on each input variable is established using a Kriging model. The non-dominated sorting genetic algorithm NSGA-II is then applied to optimize the platform's geometric scale. The specific steps are as follows: Figure 2 As shown: S51, using the established 3D design space as the search domain for decision variables, sets upper and lower bounds for each variable; the random generation scale is... The initial parent population Each individual represents a set of platform geometric scales; an algebraic counter is set. .

[0058] S52, All individuals are fed into a trained and validated Kriging agent model; the model outputs the bi-objective value for each corresponding individual—the average power generation. With power standard deviation To form the target matrix ; S53, for Perform a non-dominated sort, which will result in X non-dominated fronts (using...). , ,..., (represented by...) The parameters are obtained after performing a non-dominated sort on each individual in the population. , indicating the dominance of an individual The number of individuals in the population. Individuals are denoted as non-dominated frontiers ,remember The individual's rank value is 1; Other individuals in The rules are assigned to the next level of non-dominated front. Each iteration iterates through all individuals in the population, assigning a corresponding non-dominated front to each individual. For individuals with the same rank, crowding is calculated, and the individual with the smallest crowding coefficient in the entire solution set is removed to maintain population diversity. The crowding calculation formula is as follows:

[0059] In the formula, For individuals The degree of congestion; Number of sub-targets; for The function value; for The function value.

[0060] S54 employs a binary tournament selection method: individuals with lower non-dominant levels and higher crowding are prioritized for retention in the next generation. The simulated binary crossover SBX operator is used to encode real numbers; the calculation formula for the SBX operator is as follows:

[0061]

[0062] In the formula, , Represents the parent generation; , Represents offspring individuals; Representing algebra; Let be a random variable, in the formula let Let be a uniformly random number generated in the interval (0, 1); It is a non-negative number.

[0063] Crossover is performed using the SBX operator, and polynomial mutation is used to generate N new individuals, which form the offspring population. ; Call the Kriging proxy model again to obtain The target value.

[0064] S55, merge the parent generation and the child generation to get ;right Perform non-dominated ranking, selecting individuals with low non-dominated rank and high crowding to form the next generation of parents. Termination condition: If Output the non-dominated solution set and end the loop; otherwise, merge the parent and child generations, repeat the selection-crossover-mutation operation in S54, and generate a new child generation. And its target value was evaluated using a Kriging model; .

[0065] Example 2 As a typical example, such as Figure 3 As shown, the tension leg platform of the present invention includes a central column, side columns, buoys, diagonal braces, and a tension leg mooring system. The central column is located at the geometric center of the platform. The side columns are arranged around the central column at equal angles of 120°. The buoys are located at the bottom of the central column and the side columns, serving as connections between the side columns and the central column. The diagonal braces are located at the upper part of the central column and the side columns. The upper end of the tension leg mooring system is connected to the bottom guide cable of the side column, and the lower end is connected to the seabed anchoring foundation.

[0066] Furthermore, the central pillar is a cylindrical hollow float; the side pillars have the same radius as the central pillar and are provided with a cross-shaped inner compartment that runs through the entire height of the side pillar. The inner compartment is composed of two vertically arranged watertight bulkheads that intersect at the central axis of the side pillar, dividing the inner cavity of the side pillar into four fan-shaped ballast tanks.

[0067] Furthermore, the pontoon adopts a circular cross-section, and the connection nodes between the pontoon and the central column and the side columns are equipped with cast steel bifurcated parts to ensure smooth force transmission.

[0068] Furthermore, the diagonal brace is a circular tube component, with both ends welded to the outer walls of the central column and the side columns via node plates.

[0069] Furthermore, the tension leg mooring system includes an upper cable guide, a tension leg body, and a lower suction anchor. The cable guide is located on the outer side of the bottom of the side post and is used to guide the tension leg to form a predetermined clamp with the vertical line of the platform. The suction anchor is an inverted barrel-shaped steel structure, and the depth of insertion into the mud is determined according to the seabed soil.

[0070] This invention provides a parameter optimization design method for a tension leg wind turbine platform under fully coupled pneumatic-hydraulic-mooring operation, comprising the following steps: Step 1: In the CAD / CAE environment, create a 3D solid model of the platform, including the central column, three side columns, bottom pontoons, upper diagonal bracing, and cross-shaped inner compartment. Determine the relative positions and main dimensions of each component, and calculate the platform's mass, center of mass, moment of inertia, draft, and displacement. Mesh the wetted surface with one click and output in .gdf / .hst formats for subsequent hydrodynamic calculations.

[0071] Step 2: Calculate the added mass, radiation damping, hydrostatic stiffness, and wave excitation force in the frequency domain solver, and import them into the time domain software. Establish the anchor chain model using the lumped mass method, extract the mooring stiffness, and substitute it back into the frequency domain for correction. Then, establish an aerodynamic-hydraulic-mooring coupled model for time domain calculation, and output the platform motion, mooring tension, and power generation time history.

[0072] Step 3: Using the radius of the central column Radius of the side column Distance from the side post to the center Define the search domain for continuous variables.

[0073] Latin hypercube sampling was used to generate An initial individual, each individual This ensures uniform spatial coverage. The coupled model is applied to the sample points to obtain the target vector:

[0074] Establishing Kriging Prediction:

[0075] Wherein covariance:

[0076] Solving hyperparameters using maximum likelihood estimation , , Complete the training of the proxy model.

[0077] Step 4: Initialize the parent population Set the algebraic counter Use the Kriging model to evaluate the objective:

[0078] right Performing a fast nondominated sort yields the front sequence. , ...; For individuals of the same level, calculate the crowding degree and prioritize retaining those with higher crowding degrees.

[0079] Step 5: Generate offspring using simulated binary crossover (SBX)

[0080]

[0081]

[0082] Polynomial mutation produces new individuals, which are then merged to obtain... ;right Perform non-dominated ranking, selecting individuals with low non-dominated rank and high crowding to form the next generation of parents. Termination condition: If Output the non-dominated solution set; otherwise, output the solution set. For the parent generation, repeat the selection-crossover-mutation operation to generate new offspring. And its target value was evaluated using a Kriging model; .

[0083] Step Six: Obtain a trade-off solution from the non-dominated solution set Substituting into the coupled model, we obtain the mean and standard deviation of the power generation corresponding to this solution. If the relative error with the proxy prediction is less than 5%, then... To ultimately optimize the master scale.

[0084] The description of the structural performance improvement of the tension leg platform after NSGA-II optimization only gives a qualitative trend and does not represent all the quantitative benefits that the proxy-evolutionary coupling algorithm of this invention can obtain. The description does not list the iteration process of the decision variables "quantity" in each generation of the population, nor does it publicly display the precise scale of all Pareto front solutions, but this does not mean that this algorithm only outputs qualitative trend analysis.

[0085] Example 3 A tension leg floating platform coupled motion response and main body parameter optimization system includes: The tension leg floating platform model building module is configured to build a tension leg floating platform model based on the geometry and parameters of the tension leg floating platform. The frequency domain calculation module is configured to perform frequency domain calculations based on the tension leg floating platform model; The mooring line model construction module is configured to determine the layout of the mooring system and the mooring line material based on the tension leg floating platform design scheme, build a mooring line model based on the tension leg floating platform model, and perform static analysis on the mooring line model. The correction module is configured to correct the frequency domain calculation results based on the static analysis results; The wind turbine model building module is configured to build a wind turbine model based on the corrected frequency domain calculation results and the mooring line model, combined with the wind turbine parameters. The coupling analysis module is configured to perform aerodynamic-hydraulic-mooring coupling analysis based on the wind turbine model and the mooring line model; The optimization solution module is configured to select the key geometric parameters of the tension leg floating platform as optimization variables based on the coupling analysis results, construct a surrogate model of the objective function for each input variable using a Kriging model, and use a genetic algorithm to perform optimization and obtain the optimization results.

[0086] 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 one or more computer-usable storage media (including, but not limited to, disk storage, etc.) containing computer-usable program code. CD - ROMIt takes the form of a computer program product implemented on (such as optical memory, etc.).

[0087] 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.

[0088] 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.

[0089] 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.

[0090] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art without creative effort within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the coupled motion response and main parameters of a tension leg floating platform, characterized in that, Includes the following steps: Based on the geometry and parameters of the tension leg floating platform, a model of the tension leg floating platform is constructed. Frequency domain calculations are performed based on the tension leg floating platform model; Based on the design scheme of the tension leg floating platform, the layout of the mooring system and the mooring line material are determined. A mooring line model is established based on the tension leg floating platform model, and a static analysis is performed on the mooring line model. Based on the static analysis results, the frequency domain calculation results were corrected. Based on the corrected frequency domain calculation results and the mooring line model, and combined with the wind turbine parameters, a wind turbine model is established. Based on the wind turbine model and mooring line model, aerodynamic-hydraulic-mooring coupling analysis is performed; Based on the coupling analysis results, the key geometric parameters of the tension leg floating platform were selected as optimization variables. A Kriging model was used to construct a proxy model of the objective function for each input variable. The optimization results were obtained by using a genetic algorithm.

2. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of constructing a model of a tension leg floating platform based on its geometric configuration and parameters includes: determining the dimensions of each column, pontoon, and brace, as well as the column spacing, based on the geometric configuration and main dimensions of the tension leg floating platform; calculating the total mass, center of mass, moment of inertia, draft, and displacement of the platform; constructing a three-dimensional solid element model; and meshing the model.

3. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of frequency domain calculation based on the tension leg floating platform model includes: using a frequency domain solver, calculating the platform's six-degree-of-freedom additional mass, radiation damping, hydrostatic restoring stiffness, and unit amplitude excitation force based on the tension leg floating platform model, and further obtaining the load and displacement. The platform's dynamic response frequency domain equation of motion is: In the formula: For frequency, The Fourier transform of the platform's time-domain motion response; The system mass matrix includes the hydrodynamic added mass matrix caused by radiation. and structural mass matrix ; Here is the damping matrix; This is the stiffness matrix, including the mooring stiffness matrix at the static equilibrium position of the structure. and the hydrostatic stiffness matrix of the floating structure ; This is the environmental load matrix, including aerodynamic loads. First-order wave excitation force and second-order low-frequency wave force .

4. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of establishing a mooring line model includes: Based on the platform design, the layout of the mooring system and the materials of the mooring line are determined. The tension leg floating platform model and frequency domain calculation results are imported into the time-domain hydrodynamic analysis software, and the mooring line model is established using the lumped mass method. In the lumped mass method, the performance of the mooring line is equivalent to a nonlinear spring, consisting of several continuous, massless segments and nodes at the midpoints of each segment. Each segment is a continuous, massless pipeline element, and its axial, bending, and torsional characteristics are considered only. It is simulated as a combination of axial, bending, and torsional spring dampers. The nodes concentrate half the mass of two adjacent segments, and all the distributed forces acting on the pipeline element are concentrated at its concentrated mass point. The effective tension on the pipeline is obtained by stress-strain relationship according to the transient position of the node.

5. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, Based on the static analysis results, the process of correcting the frequency domain calculation results includes: performing static analysis on the platform and mooring system to obtain the platform's mooring stiffness matrix, which is used to characterize the platform's restoring force response under six degrees of freedom and small displacements; using the derived stiffness matrix as the external constraint of the frequency domain solver; repeating the frequency domain analysis steps; and correcting the frequency domain calculation results.

6. The method for optimizing the coupled motion response and main body parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of establishing a wind turbine model includes establishing a wind turbine model based on parameters such as rotor radius, airfoil lift and drag coefficient, blade mass distribution, and tower height, according to the corrected calculation results and mooring line model. Among them, the wind turbine blades are modeled using blade element momentum theory, the tower is modeled using the lumped mass method, and the nacelle and hub are modeled as six-degree-of-freedom solids.

7. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of performing aerodynamic-hydraulic-mooring coupling analysis includes: calculating the aerodynamic load of the wind turbine using blade element momentum theory, and the lift coefficient... and drag coefficient The expression is as follows: In the formula and For the projections of lift and drag onto the z-axis and x-axis, Indicates the relative velocity of the incoming flow. The air density is given; the lift coefficient and drag coefficient are related to the airfoil, Reynolds number, and angle of attack; when the lift coefficient and drag coefficient are known, the lift and drag on each airfoil are solved by airfoil parameters, and the load is obtained by integrating over all airfoils of the entire blade; during the solution process, an external link library is used to control the blade pitch and rotor speed, the time-domain analysis software outputs the blade position information and aerodynamic loads, and the external link library updates the blade pitch and rotor speed; The platform's hydrodynamic response is calculated using a hybrid method combining three-dimensional potential flow theory and the Morison equation. For the main structure of the platform larger than the set size, diffraction / radiation theory is used, and the time domain solution is performed based on the corrected frequency domain hydrodynamic coefficients. For structures smaller than the set size, the Morison equation is used for calculation.

8. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, A Kriging model was used to construct a proxy model for the objective function on each input variable. The process included selecting the column radius, column spacing, and buoy radius as optimization variables. The Kriging model consists of a multinomial regression term describing the overall trend of the response and a local bias term, and its calculation formula is as follows: In the formula, , Regression models and normally distributed Gaussian random processes; To optimize variables; These are the regression coefficients; The covariance expression is as follows: In the formula, To indicate and The related functions are expressed as follows: In the formula, Spatial dimension; For the relevant function value to follow The rate of change; For the smoothness of the model; The relevant undetermined parameters of the Kriging model were determined through maximum likelihood estimation, and the unknown points... The best linear unbiased estimate is: In the formula, for Least squares estimate; This is the relationship vector between unknown and known points; for The mean squared error; This is the estimated value of the target by the proxy model; Latin hypercube sampling is used to generate initial sample points in the three-dimensional design space to ensure that the samples are uniformly distributed throughout the domain and avoid clustering. Each sample point is then substituted into the aerodynamic-hydraulic-mooring coupling analysis of the floating wind turbine to calculate the platform's six-degree-of-freedom motion response, mooring tension extremes, and power generation time history. The mean and standard deviation of the power are extracted to form a sample library. Based on the base sample library, a surrogate model for each objective function on each input variable is established using the Kriging model.

9. The method for optimizing the coupled motion response and main parameters of a tension leg floating platform as described in claim 1, characterized in that, The process of using genetic algorithms to optimize solutions includes: Using the established 3D design space as the search domain for decision variables, upper and lower bounds are set for each variable; the random generation scale is... The initial parent population Each individual represents a set of platform geometric scales; an algebraic counter is set. ; Will All individuals are fed into a trained and validated Kriging surrogate model; the output for each individual is a bi-objective value: the mean power generation. With power standard deviation To form the target matrix ; right Perform a non-dominated sort, which yields X non-dominated fronts. , ,..., This indicates that the parameters are obtained after performing a non-dominated sort on each individual in the population. , indicating the dominance of an individual The number of individuals in the population Individuals are denoted as non-dominated frontiers ,remember The individual's rank value is 1; Other individuals in The rules are assigned to the next level of non-dominated front. In each iteration, all individuals in the population are traversed, and a corresponding non-dominated front is assigned to each individual. The crowding degree is calculated among individuals with the same rank value. The individual with the smallest crowding degree coefficient in the entire solution set is deleted to maintain population diversity. A binary tournament selection method is adopted: individuals with low non-dominance level and high crowding are preferentially retained to the next generation, and the real numbers are encoded using the simulated binary crossover SBX operator. Crossover is performed using the SBX operator, and polynomial mutation is used to generate N new individuals, which form the offspring population. ; Call the Kriging proxy model again to obtain Target value; By merging the parent and child generations, we obtain ;right Perform non-dominated ranking, selecting individuals with low non-dominated rank and high crowding to form the next generation of parents. Termination condition: If Output the non-dominated solution set and end the loop; otherwise, merge the parent and child generations, repeat the selection-crossover-mutation operation in S54, and generate a new child generation. And its target value was evaluated using a Kriging model; .

10. A system for coupling motion response and main body parameter optimization of a tension leg floating platform, characterized in that, include: The tension leg floating platform model building module is configured to build a tension leg floating platform model based on the geometry and parameters of the tension leg floating platform. The frequency domain calculation module is configured to perform frequency domain calculations based on the tension leg floating platform model; The mooring line model construction module is configured to determine the layout of the mooring system and the mooring line material based on the tension leg floating platform design scheme, establish a mooring line model based on the tension leg floating platform model, and perform static analysis on the mooring line model. The correction module is configured to correct the frequency domain calculation results based on the static analysis results; The wind turbine model building module is configured to build a wind turbine model based on the corrected frequency domain calculation results and the mooring line model, combined with the wind turbine parameters. The coupling analysis module is configured to perform aerodynamic-hydraulic-mooring coupling analysis based on the wind turbine model and the mooring line model; The optimization solution module is configured to select the key geometric parameters of the tension leg floating platform as optimization variables based on the coupling analysis results, construct a surrogate model of the objective function for each input variable using a Kriging model, and use a genetic algorithm to perform optimization and obtain the optimization results.

Images